Vinicius Ferreira: Morning.
george wang: Hey. Good morning, everybody.
Lucas Menegazzo: Guys.
Vinicius Ferreira: Okay.
Lucas Menegazzo: Looking good,
David Kawata: Good afternoon.
Lucas Menegazzo: David. On-It it it just raised the bar
george wang: David got some Alright.
Lucas Menegazzo: for us.
David Kawata: Well, I I knew, mister Wang was gonna be on the call, so I still don't wanna wanna bring it. Day one. Let's go.
george wang: It's exciting. Okay. Let me pull up what we prepared in advance of this. K. Just nd all my all my screens. Alright. So we put together can you see? Alright. We just put together things on the slides because that way it's a little bit more linear to walk through. There's a lot of information. That we digested from the notion, from, you know, everything.
george wang: And, yeah. And even, Yeah. So yeah. Appreciate, again, appreciate you guys for making the time. So, yeah, the team spent this weekend adjusting everything, and, basically, this is an artifact to help us align and The thinking behind what we're gonna build. So we're gonna be talking through ve things today, what we heard, making sure that we got the foundations right, some insights, three insights from the research that, you know, what we looked at the monster leads group's website, the the blog, and to, you know, figure out what what are the what are the opportunities that, that we should identify.
george wang: Really figure out what calm confidence means to us as a direct design direction, and then the core journey. And then nally, we're gonna be talking about next steps for this week. Okay. So before we show you what we're thinking, we wanted to make sure we got the foundations right. So these four principles stood out to us from the kicko . Clarity is our goal. It's achieved when the user understands what matters, knows what does not, and feels safe pausing.
george wang: We want to orient them, not obligate them to convert. We wanna give before we ask. It's not a door with a lock, and we wanna always make sure that ask is smaller than to give. And lastly, we want to let the user drive. It's a treasure hunt like they're going to the rotisserie chicken. Oh, and we wanna meet them where they are, and everything is built around these four pillars. Okay. So we're gonna be, sharing some of these sites.
Lucas Menegazzo: Sure.
george wang: Sorry. Did you guys say something?
Lucas Menegazzo: Just to point out, if you can go back, one screen just so I can highlight something here.
george wang: Yeah.
Lucas Menegazzo: The orientation, not obligation is also a legal requirement. We cannot, recommend something. We can orient the user to do so. So this this should be in mind as well.
george wang: Right. Okay. Sounds good. Note that down. Okay. So three insights from the research that shape how we are approaching the product. Then yeah. So this is, insight number one. Our wedge is doing what existing vendors aren't paying attention to. So we arrived at this by reading the blog of Monster Leads Group, given Ken is the owner, and we found this graphic.
george wang: Five reasons that people refinance. And while equity access and payment reduction are factors driven by rates, I need money or I need to pay less, the term reduction is the one motivation that most lenders, because they're driven by desire for freedom rather than needing money or paying less at this moment. Yeah. That's just, you know, something that, the lenders are not paying attention to, you know, because they're So focused on the campaigns that oriented around rates.
george wang: This is an opportunity that we have to, do something unique, relative to the market. And we also said that saw that 21% of the qualified market actually have multiple well, 27% have multiple, motivations, and that means we should position the product around these, blind spots. By really, you know, presenting all of the possible solutions to the, the borrower.
george wang: And, yeah, so that's the how we could really, position ourselves entering the market. The segment insight is behavior shaped by psychology, not just math. We came across Two behavioral economic studies on mortgage refinancing. On the left side, researchers found that the complexity and the effort that it takes to refinance is really what, you know, holds back a lot of families from doing so.
george wang: And this validates the product hypothesis that we need to make it, you know, as simple as possible and, really guide the borrower. And on the right side, the city of Chicago, I think they sent out this, they did this mailer campaign, offering, preapproved zero cost refinance, and 84% didn't respond and only very few people did.
george wang: And what they the interesting that they found is households are 7% more likely to refinance if a, neighbor within 50 meters has recently refinanced. So that means that social that we you know, it's something that we didn't consider before. It's something that we by bring embedding that into the product, we can encourage people to refinance more. Somebody raise their hand.
george wang: Hey, David.
David Kawata: Yeah. Yeah. So in in the write up, there there was a reference from a book called influence, and they did reference the yeah.
george wang: In uence.
David Kawata: So it talks about social social, proof, social, validation.
george wang: Yeah.
David Kawata: So it's actually already in the write up. And so, it's actually there. And then go ahead. Thank you.
george wang: Sounds good. Okay. And news site number three is that, MLG data can help us meet borrowers earlier and more intelligently. So we try to look at the thing. What comes before and after it is, you know, important for us to, you know, really hone in on what ClarityEngine should be. And in the kicko call, we talked about driving users to ClarityEngine through influencers and then, in the future, integrate ClarityEngine into real estate CRMs.
george wang: But what exactly is the kind of content that we could offerinfluencers and what is the value of the CRM integration? And we got some of the, you know, insights from reading the blog. No content pointing to what other lenders are not doing. This could be a content strategy that we could equip the influencers with. And for the Clarity Engine, we came across the stat that across, active campaigns, the average loan amount for borrowers pursuing this type of refinance, a conventional rate in term finance, is approximately 425.
george wang: So, You know, based on the previous insight that refinance behavior is largely influenced by neighbors, I think there's an opportunity to frame other these kinds of MLG data in a way that tells borrowers what others like them are already doing. So I wonder, you know, what other data points can we get that says people like you are doing, you know, x y z to help them to nudge them to, to, you know, get to clarity faster.
george wang: And on the b two b two c side, the emoji blog also have the importance of equipping, loan officers with tools to be able to explain, certain topics, to borrowers. And, when all of us get these today, they barely know anything about them. So getting that clarity from the Clarity Engine and turning that into playbooks for the CRMs, that's something that additional value that it's kind of that's additional value that the Clarity engine can o er. So Before and after really helps us hone in on what clear how the design clarity engine.
george wang: So that's the research. Now going into the design direction for calm confidence and you know, we looked at so this idea of confidence, calm confidence, it, starts with a sense of place. So when you enter your address, the product should treat Street's a place. It's grounded. It's spatial, not just an abstract asset, number. Right? Your home, your street, your neighborhood. This is all, you know, a reference or how property data can feel personal without feeling like listing.
george wang: Con dence is serious, clear, and not intimidating. Fifinancial information that, you know, doesn't threaten, like, Apple Wallet, Rocket Money, and Way that's calm, hierarchical, and honest. You know, rug and money here, it it gami es the experience of managing your personal nance. And, you know, this is how scenario cards should feel, serious enough to trust and clear enough to understand that they're not intimidating.
george wang: The third is, you know, David earlier emphasized the need to hold the full picture. Accounts, loans, investment, all in one view. And that's energy for your mortgage in context, right, not just your rate, but, you know, your equity, your neighborhood, and, all your options in your picture. And calm confidence means that, you know, there's room to explore, pause, and come back. The product remembers where you are.
george wang: You can leave and come back like, Airbnb's trips tab. Your current context, Net ix is continue watching. Duolingo is you came back, and Credit Karma is your score went up, after you left it. These are return states, patterns that we can borrow. And for pre le, that means the scenario is saved. Come back anytime. We'll update the picture when things change. And, lastly, we wanna allow users to compare real options without pressure.
george wang: Maps allow you to pick different forms of transport and, you know, shows you different, potential routes to get to where you wanna go. NerdWallet allows you to select multiple cards and compare them. And Google Flights shows the date grid and, you know, different options and all in one view, and Redfincontextualizes the 29% debt to income ratio, as comfortable. And, the emotional tone of the product is it's not, you know, mortgage ads.
george wang: It's, daylight, wood, openness, homes that feel real. This is the world that the product lives in, and calm confidence isn't cold. It's not minimalistic. It's warm, grounded, and human. Any questions before we move on? K. So going through the journey, before I go through this. So this is, I'm gonna explore this in Figma later on, but we divide the journey into five steps, land, discover when address is entered, converse with, home advisor, explore scenarios and different options, and get into clarity of next steps.
george wang: And how we've, you know, visualized it is the strategy that we're referencing on top and what they see and what they do and references of products that do Of those things well. So, we're just gonna go through the stages 1 x 1. So remember, insight number two, eighty four percent of people ignore free, refinance o er. The barrier is not awareness, but it's avoidance.
george wang: So there's three three strategies here. First, it's, to overcome psychological barriers by showing how they compare against their neighbors. 7%, more likely what a neighbor recently did, and that's that's something that we should pay attention to. Overcome privacy. You know, start yeah.
David Kawata: Hey, George. Real real real quick. That's actually not true.
george wang: Yeah.
David Kawata: So I I I wanna the awareness so you you own property in Toronto. Yes? A busy guy. You have a pretty probably high hourly rate. There is a minimum dollar amount that makes refinance make sense for you, and that dollar amount will probably be different than Lucas and Vin and Vinny and myself. Right? So it's not necessarily always awareness.
george wang: Mhmm.
David Kawata: It's also is the squeeze worth is the juice worth the squeeze? Like, refinance is painful. People don't wanna go through the process. So just wanna wanna be kinda clear. The awareness That that's then worth it to go through the process. Like, if you save a $100 a month, a $100 a month for thirty years or three hundred sixty months, that is actually real money. Right? But in reality, a $100 a month is maybe barely the minimum that people will talk to on saving because it's so painful.
george wang: Right.
David Kawata: Just just want a slight adjustment on the awareness is not just, can I refinance? Is it gonna be worth my time to refinance? And that is a moving scale for each person on this call.
george wang: Oh, that's absolutely true.
David Kawata: In each household. So I just wanna be clear.
<v george wang>And, you know,
David Kawata: The benefit has to hit really quickly to solve
george wang: That's something that's come up, you know, over and over.
David Kawata: the problem that's then worth the process.
george wang: I think there is something to be said about somebody like you refinance, and this is the situation that they're in because it's really hard for somebody to, you know, to to to like, there's this, you know, pressure to there's a lot of pressure. There's a lot of, you know, effort and, you know, numbers you have to look at. And, you know, by showing you, Who who, like you, have done something similar. It might just give you the motivation to, to go through what you think is might be a lot of, you know, pain and effort, but really isn't.
george wang: I mean, that that's just the the point we're trying to make here. Yeah. How do we incorporate that social factor into the the opening? Into, you know, when you land on the experience.
David Kawata: That that's ne. And I agree. The social proof for influence on creating an awareness, a likability to say, okay.
george wang: Mhmm.
David Kawata: They know me or they know my people. I get that. I just wanna be clear. You talked about an awareness factor on the zero point zero cost. That is in relationship On-It shows to a percentage of value because each Did not refinance or purchase their home in the same time Roam. Same purchase price, same time frame. So interest rates and variables are different,
george wang: Yeah.
David Kawata: meaning their benefits are different. I just wanna be clear that, yes, that, my peers, my neighbors,
george wang: Right.
David Kawata: this is great, and the but to have a real clear of awareness of my benefit that's gonna be worth me listening is actually
george wang: Mhmm.
David Kawata: the that we have to hit faster and more upfront. Because it's all about what's in for me, not it's not a social proof or tribal being awareness as as that that's important that we we can meet their needs. There's likability, there's trust, But at end of the day, it's about net tangible benefits. Just wanna be clear that that is the on the presentation for the
george wang: Okay. Sounds good. I'll make note of that.
David Kawata: awareness that, I I I didn't hear that from
george wang: Yeah. I mean, we all be we always been in you know,
David Kawata: your original presentation.
george wang: we all been in this actually,
David Kawata: So thank you.
george wang: I have I have been in a situation where I put my, name and email down on, you know, I think it was LendingTree. I just get so much Roam. And so being focusing on not, you know, sharing your phone number with anyone and being really, fun and explicit about that, is important. And overcoming the avoidance by reducing the stake by making it safe to just see where you stand. Patagonia does does really well. They say don't buy your jacket. Well, don't buy this jacket.
george wang: We can do something like, you might not need to refinance. Let's nd out. And that that may just help over you know, your point considered this, in addition, could be a way to message it so that, you know, it help helps you overcome that. Perceived effort to just sometimes that Yeah.
David Kawata: Right. So in in that scenario, it's like, well, mostly your neighbors won't refinance for under $250 a month. Why don't we see how much we could save you? Right? To that concept, social proof, maybe payment, restructure it, but I wanna be clear. It is it is seems to be consistent on the awareness. It's not social proof. It's how much it's gonna benefit me and if that meets my hurdle rate. So I just To this point, don't buy my neighbors. Right? If we could create a threshold of what we've helped in the neighborhood or some way where we do the takeaway, if they don't do this, unless we save you x.
David Kawata: Right? But it's it's to that awareness. I need to put money in my pocket. Now with that also being said, there are other reasons it's gonna save me money. And so I wanna make sure that when we if we're leading with that, we don't limit the other normal benefits. So that that's a challenge until we get to the why. Right. Why are you here? And now it's just what it was stated. One of the one of the key things that we're trying to make sure that we do is not the stated goal from the consumer, but it's the real why that they've asked.
David Kawata: Right? Because a lot of times people wanna save money. Everyone here wants to save money. That that's a given. How they save money in unsecured credit cards, student loans, payments, saving money so they could save for college tuition. Right? The saving money generally is the unlock or bigger why. So I wanna just articulate this in the document. That that that is gonna be our challenges from a linear, a linear, don't buy this jacket unless we've already clarify from the consumer what their defined purpose is.
george wang: Mhmm. Right. Okay.
David Kawata: Then we could do that take away negative
george wang: Makes sense.
David Kawata: negative neg close.
george wang: Okay. So
David Kawata: So I just wanted that would be a just a detail.
george wang: once they enter their address, the the the goal of that first,
David Kawata: So thank you.
george wang: step of the journey is just to get them to enter their address. And the second, after they after that happens, this is where the the Costco metaphor comes in. Like, it's it's like a treasure hunt. Yeah. Every bead is a discovery. It's not a Roam. You know, we wanna, again, just leverage the neighbor effect with real data showing, you know, how much their entirefineighborhood could save, if they refinance. We wanna break the, cut the inertia by, just having the math done on the address, pulling all the data from Adam and all the Data sources and be able to show them, you know, this is how much they could save.
george wang: And so what they see is a progressive reveal, using aerial, like, Roam view of their property, using Google's, new feature. We wanna make it emotional, personal. This is your Roam, and, you know, the property data phase in, the value of purchase date, the neighborhood trend, then the refinance date estimated equity. Approximate remaining balance. And then, you know, the camera pulls back and be able to kind of show that you versus your neighbors kind of or you versus people like you that may not necessarily be in your neighborhood.
<v george wang>Here's what you're looking at, the benefit they could derive. And all this is, based on the one address input that they do they enter. And so And, there's some here's the you know?
David Kawata: So one of the other things that we need to do is making sure that it's uid. So if the customer does not enter their address, this is not a just a hard no, hard door. Right? So we need to be able to, right, make sure that we give to get to that value to get to the or into the address. So I just wanna see, do you have a plan for that as well? Right? So we enter,
george wang: They don't enter their address,
David Kawata: we do the call to action, wherever the process is. If we don't get them to land into
george wang: I mean, well, think about that.
David Kawata: this
george wang: Think about how to you know, what are the different routes that they could take
David Kawata: So so proof point,
george wang: to continue to keep them engaged.
David Kawata: until we've earned that,
george wang: Maybe, you know,
David Kawata: do you have a plan beyond is
george wang: there maybe there's something that it where they can explore,
David Kawata: it would it be just text? Is that would be your thought?
george wang: for, educational resources or, friends in refinancing. You know, learn what they learn about things that About refinancing and, you know, what, what they could potentially qualify for, what they could potentially nd out once they enter their address.
<v David Kawata>Right. Right. But I will again, on on the brief, I wanted to make sure that you didn't articulate that we need to be even positioned to give value
george wang: That's something that we'll we'll do back in,
David Kawata: even if the consumer did not opt into the property
george wang: and continue to Mhmm.
David Kawata: address from step one, that there is another pathway that we have the uidity. I did not hear that in the the your presentation, your tone, that is understood that it's not a hard stop if we don't go from step one property address. No. I appreciate that. But but that that that was in the brie ng. It's actually a kind of an important concept that there's that all the doors that we have are two ways. Right?
David Kawata: And and there's there's no walls. So we're we're continually owing. To that point, if a consumer wants to just ask general questions about what the general rate is in my area or what are the options on beyond lowering payments or how do I harvest equity
george wang: Understood.
David Kawata: or or that kind of options?
george wang: K.
David Kawata: Calculators,
george wang: Yeah. So in terms of the conversation
David Kawata: terms, and the ability to provide value.
george wang: with a home advisor,
David Kawata: The the I wanna make sure you read the brief
george wang: first, we wanna, you know, make the absence of a speech feel like
David Kawata: in that context because that that that is a way
george wang: a feature. So and the system never recommend,
David Kawata: for us to still give value in a conversation without getting something
george wang: never rank, and never says you should.
David Kawata: specific like an address.
george wang: And, second, it's adopting the conversations, the borrower's mode. The Lanes.
David Kawata: Yes. Thank you.
george wang: We detect the archetype of the borrower and the preferences through the conversation that we adopt, the tone, and, how like, the the way that the AI engages with them in real time and get smarter. So what you see is, you know, what's on your mind about your home. It's a free text. It's not a multiple choice. And it's, living in at the same time, they see a living profile that reflects back what the system understands about them, which updates as they share more information, which, you know, they can correct at any time.
george wang: And that's no that's how they know they've been heard if they see what the system, you know, like, summarizes about the conversation so far. And if the borrower receives something heavy, And acknowledges before solving it. Saying something like it sounds like a lot is changing right now, kind of like this, therapy app. In reference here, when the user writes something, there's an acknowledgment before moving on in the conversation ow.
george wang: And in the explore stage, the Borrower AI helps, translate what the borrower, is experiencing in their life into refinance options. Someone says, I'm going through a divorce. I need to figure out the house and system translating that into equity access and possibly combined with payment reduction. And it probes for motivations without attening the complexity that they could all exist since twenty seven percent qualify for multiple programs.
george wang: The system holds all of it and doesn't force them into one box. What they see is basically scenario cards organized by their goals, not scenario a, nish faster, lower your payment, access your equity, stabilize my rate, cover your equity, and each card shows what you gain, what it costs, new payment versus current, and the total interest breakeven. And the payo date.
george wang: And every card has a disclosure, don't rank, and we don't have it a winner. We let the user, depending on the archetype that they're in. If we're analytical, risk taking archetype, then we allow them to, tweak the sliders in order to see, the trade offs in real time. If it's a you know, they're the practical cautious archetype, then we allow them to You have to say, these are two scenarios that I like. Can you just combine them and give me the answer?
george wang: The scenario that is the the best of both worlds. And the emotional archetype, the AI is, you know, ask you know, basically guiding you to tell them more details and interprets the refinancing constraints based on, your needs. And here, obviously, it's a visual of how, well, front path handles the, interaction of the optimization scenarios.
george wang: You know, interactively figure out, visualize, what the potential paths could be. And lastly, in the clarity stage, We create a ywheel by having users share their refinance information with others. And borrowers who share that, you know, can manufacture the neighborhood effect at scale. Second, refinancing is a decision in a relationship On-It by themselves. So whether it's partners, families, advisors, and the product should facilitate that agreement, right, and not just inform that one person.
george wang: And the thirdly, Yeah. The user can invite others and, save their progress and and alert it by leaving their email, and this is a give, to get principal. So that's something like wanna keep, we built, at the end of the email, and it saves it. And, potentially, there's an option to alert them when something when their rates drop or there's an opportunity to To, you know, get a better scenario.
george wang: Yeah. I'm inspired by, you know, Spotify. It's called yeah. Spotify wrapped. This is drawable achievements. So anything that, could summarize what their the the the scenarios that they landed on. That they could share and come back to, if they want in the future. So that's the that's the user journey.
george wang: Any questions, comments?
David Kawata: One one of the things on the shareable is actually key for call it the UI, the design is called the mortgage readiness report. Right? And there may be different, expandability of that depending on how much information the consumer has validated. So I wanted to just highlight the share. Again, I I I don't think it's defined what would be in that report or would be into that shareable for the reality, but I wanted to highlight. I think that's a really And virality, for social proo ng of their trusted partners,
george wang: Right. Okay.
David Kawata: and then making it actionable.
george wang: Yeah. So, you know, for next steps,
David Kawata: So,
george wang: we're here day one of week one.
David Kawata: we would like to make sure that we think about what are the contacts
george wang: We're gonna have daily check ins for the rest of the week and by the end of the week.
David Kawata: in the report that is, one, in a prequali cation. I can give you an example of the information in a prequal, because but this is not a credit o er. So a mortgage readiness will be a variations of that. But I think that's a really great Cool. That for a UI, we need to, put that into sub something that we should have an idea what that looks like. So thank you.
george wang: That yeah. I took a look. There are some, you know, stars, and I think the team had gone to, you know, star of the ones that the stickies that, you know, stood out.
David Kawata: Actually, real quick.
george wang: But,
David Kawata: Vint, Vint, you have different stages in the Vinny,
george wang: I saw that was organized by different archetypes, or different motivations.
David Kawata: do you have any any thoughts about,
george wang: And There seemed to be a lot of variation between the different journeys.
David Kawata: George's, you know, so we can create a dialogue?
george wang: So, yeah,
David Kawata: But with that with that being said,
george wang: I'm kinda curious to have,
David Kawata: George, did did you get a chance to take a look at,
george wang: you know, Vinay or the team walk through some of the,
<v David Kawata>any Vinny's and and how it's slightly
george wang: you know, some of that and and,
David Kawata: different as far as the naming conventions.
george wang: contrast compare with what what have we shared.
Vinicius Ferreira: So oh, okay. So there's a lot of, things that are, converging. Yeah. So, on my user ow, the user journeys, yeah. So, I tried to understand, like, different scenarios. And, okay. So firstthing first, I like that you brought some references. Couple of things that they're okay there.
Vinicius Ferreira: There's something that, there are di erences for for this context. But, for example, I think this is more like a an, a conversional a a UI instead of, like, a step like, a wizard or something like this, like, a step by step. Is more uid. Also, the the finalreport,
george wang: Wanna pull it up and we can take a look together?
Vinicius Ferreira: I believe that is is going to be I don't know why it's, like a PDF or something like this, but we need to figure out that. But there's a lot of things. Like, I like the calculator that you brought. They're super nice. I like that you brought that, people can visualize the neighborhood. This is also something, interesting that, but okay. And oh my gosh. I forgot my my, my thought, line here.
Vinicius Ferreira: But, yeah, I I I think
David Kawata: So how about how about how about we do this
Lucas Menegazzo: We we we Yeah.
Vinicius Ferreira: that Yeah.
David Kawata: for time saver agenda? I'm gonna let Lucas talk. But why don't we do this? Why don't we let George nish the presentation? So, George, I apologize for disrupting. I I wanna make sure you have your ow. We get that. And then, Benny, why don't you just take a moment after you see it, and then we can converse with notes. That way, we're at the same point. With that being said, George, I think you had a you looked at Benny's, but you didn't look at Benny's. And Benny hasn't looked at yours. Let's just kinda Finish. So you haven't so you you've prepared a lot of, time, and I I wanna respect that for our agenda time.
<v David Kawata>And, Lucas, if you have a question first, and then we'll get back to George. And then, Vinny, we'll let you circle back later. Is that okay?
Lucas Menegazzo: Just a quick side note here. For the conversation side, I'll I would like to to, you to take a really good care and focus on that because conversational tag will be also very critical in terms of sending out that common confidence, wording as well. I know you focused a lot on the visual side, but this is very conversational at the the very end. Wordings,
george wang: Yeah. Sounds good.
Lucas Menegazzo: phrases,
george wang: Yeah. We definitely,
Lucas Menegazzo: ways to bring that,
george wang: you know, align, connect, and,
Lucas Menegazzo: feeling as well in terms of writing,
george wang: discuss where the, you know, where the,
Lucas Menegazzo: will be very good to to
george wang: you know, things that are aligned and diverging.
Lucas Menegazzo: see while
george wang: And, in terms of the,
Lucas Menegazzo: you're developing. And we're just reinforcing that Vini has made some really good progress
george wang: the wording and the phrasing for the,
Lucas Menegazzo: on the customer journey. There there's some,
george wang: HomeAdvisor AI,
Lucas Menegazzo: some additions to what you brought. I think,
george wang: since it's not going to be,
Lucas Menegazzo: we should just see each other's work and,
george wang: you know like, there's a lot of things that's already documented as to how
Lucas Menegazzo: reconnect tomorrow just to Ask questions and and align what we,
george wang: And there's and it's adaptive as well.
Lucas Menegazzo: we can bring together what are the di erences
george wang: So it it's it's more about what kinds of prompts
Lucas Menegazzo: and just to to align both works.
george wang: do you,
Lucas Menegazzo: So don't we we don't have any of your works to do,
george wang: provide to the AI and then,
Lucas Menegazzo: but so far so good.
george wang: you know, testing it and seeing whether it's responding tofficertain, personas in the way that it should. So I think that's the approach that should be taken to, you know, test the the personality, if you will, of the and responsiveness of the This is done. But, yeah, I was just gonna nish by saying this is the the goal of the week to you know, for us to de ne the journey. And by the end of the week, we're, we get to the alignment of, you know, the architecture so that the big Roam can, look at the feasibility, investigate the feasibility, and then we'll come back in week three.
george wang: And, that's when we have the That's when we're aligned on the, the the design system and the components, and week four is just cleaning up the the finalhandoff. So that's, I just wanted to preview that. That's about it. That's the the end of the presentation. Yeah. We'll we'll reconnect on, Roam and, figure out, perhaps the rest of the day.
george wang: How we can get more aligned, from this one. But appreciate, appreciate you guys' time. And, yeah, looking forward to moving this further. The Sigma. Yeah. I can send it to PDF or I can drop it. I have I can put in the project.
george wang: I don't know if I was invited. I tried to access Figma earlier, but seems like I had to request access, which I can. But I'll drop it into segments. I'll work with that. Mhmm. Okay.
Sydney Bocik: Will you be sending this, over by email or drop it in the chat with Roam, a link to it? This PDF? Or will you be able to move this into another project that we have?
george wang: Not for now. But if I need anything, I'll I'll ping you guys in the room. Figue.
Sydney Bocik: Okay. I can I can make sure the access to I can make sure you have the access you need? Sure.
David Kawata: So next text, George, you said that, you would we have a meeting tomorrow, Lucas, since we say that?
Lucas Menegazzo: Yeah. We do we do have dailies with with them, at the same time.
David Kawata: Alright. George, is there anything that you need, anything additional besides access to Figma? Got it. And then, Vinny, I guess he'll provide this That's where? Your your press okay. You're gonna put that in Figma as well? Okay. Great. Chris, is there or Christopher, excuse me. Is there anything else that you want to, to to add?
Christopher Everett: There's sorry, guys. There's there's not a lot of time that we have left, but, this is a a great start to it. You guys will see it, in the next one. But I love the points that you, made, David. Thanks for just, you know, really throwing that On-It was a it was a crunch this weekend, but, it's good to point out those things. We're gonna Yeah.
Christopher Everett: We'll keep going. And we gotta we got to, meet with Minnie too, so I'm good.
David Kawata: Great. Perfect. And and, again, and I and I like the diversion concept so we have a conversation. Right? So, again,
Christopher Everett: Yes.
David Kawata: I I I like I like the uidity of the conversation, but with whether the components, the calculators, the visualization, the neighborhoods, right, those are the moments, the magical moments, whatever. So, George, I really love that you brought in the best practices and other examples that you bring in that's in the native right now. And, so that we I could best merge. So I appreciate that. So, again, I wanna be respect everyone's schedule, so I wanna give you back some of the time. So, thank you, Sid. I'll turn it over to you to close.
Sydney Bocik: Access to Figma, you have full access now, and I will send out the notes from this call so everybody has full, visibility to all the points. And if you need anything else in the meantime, let us know. We're excited to see how, we can, bring the work together and kinda move forward from here.
george wang: Yep.
Sydney Bocik: So we appreciate all the work you guys did this weekend. Know there was a lot of documentation. So, I like the visuals, and it'll be I think it'll be really exciting to see how we can marry,
george wang: Great.
Sydney Bocik: you know, because most most systems are especially like Claude, Chai GPT, like, they're all conversational. So I think the visual components you brought in,
Lucas Menegazzo: Mhmm.
Sydney Bocik: it'll be a really exciting experience to figure out how we can marry chat and visuals where I think a lot of systems are which is really the next evolution of where These LLMs, I think, are probably going. So it's it's an exciting position to be in to kind of explore that. So thank thank you, guys.
David Kawata: And one of the things I'm excited about is that is that in chat today, the visuals that are coming out are in that kind of ow, kinda like what you showed in a chat in the chat conversation. Like, that was so 20 '5. But it'd be interesting because if we have more screen,
george wang: Awesome. Yeah.
David Kawata: how can we create that more immersive?
george wang: Take care.
<v David Kawata>This is where I love the Christopher Everett and the Everett and Nicole concept where they could break that fourth wall if we look at a movie concept. Like, where where can we expand? That that social proof that it's not, obvious, but is subconscious. Like so I think it'd be an interesting way where we could see, alright. Here's the linear conversation, but we have this other white space. Can we use this for the same points that you brought in? Can we say, hey. Your neighbors on average say x.
David Kawata: Would this be, like, all those type of things you brought up? So it'd be, fun to explore, and I appreciate your your guys' time today.
Sydney Bocik: Thank you, guys.
Sydney Bocik: Please continue. Thank you.
George Wang: That's a very pleasing, intro. Other than Zooms. So there's the inner information architecture. What are the big areas of this app, you know, on a desktop, and how are they going to, how are they gonna, work with each other on mobile, and what is the job that how how do we, you know, organize the components, the content, you know, interactions? Amongst these, you know, different zones, these areas, these zones.
George Wang: So this is, something we're just exploring. So when you enter, there's the there's the chat. It's just open ended question and then enter address, if, yeah, as a next step if you don't respond. And then once you enter the address, it populates this That's where the aerial view comes in, profile about you.
George Wang: The details here, I wouldn't necessarily, you know, anchor on too much. The point of what I'm showing you now is just to visualize how the different what are the different areas of the screen and what's the job that they're they're gonna do. Right? And then the scenarios, these are really the Mhmm. The prize. Right? At the end, you're To the back and forth and through, you know, us reflecting that we know what your house looks like, you know, what your how much mortgage you still owe.
George Wang: We, you know, we we come up with these scenarios for you, and that's that takes the majority of the screen as real estate. And so that's just the information architecture. Anybody have questions? I'm gonna check my room. There's it. Hands raised. Okay. Alright. So there's that. There's there's the magic moments, right, in in that that we need to create. And this is just the first step. Right? Like, there's, there's things that we want to educate the user on, the financial, refinance component, concepts, amortization, compound interest, you know, breakeven point.
George Wang: Trade o s between these, you know, Roam length and monthly payment and so forth. And whatever it is that we, that triggers the magical moment, whether it's confusion ff ff or, you know, that that's something that, another piece that, you know, we we need to kind of grab the, grab the requirements and try to prototype that out as well. But when that gets triggered Then you see something like this, something that's more interactive, that's more, you know, just the information that's displayed in a way that's easier to understand.
George Wang: So this is, like, pick two. Right? You can have two of these lowest payment, fastest payo versus this, you know, monthly pay these are very, like, concept oriented, and the magic moment of, explaining that concept to you is Much more human. Say compound interest. This is specs, of the of how compound interest works, and this is something that's easier to understand.
George Wang: For every $1, you borrow on a fteen year Roam. You pay back loan and the interest. Right? So we did that for the breakeven and the amortization as well. And another thing that Vinnie's, journey map really helped accomplish is it's, it's hard for me to switch back and forth with Figma, but there's a magic moment step at every for every journey.
George Wang: And, basically, the job that we need to do in those moments is for, each of the ve journeys is The the sticker shock. Like, when they see some when they see the the the the penalty of, refinancing, like, and they're not reacting, this helps them ease into that idea. Right?
George Wang: That's dependent on the knowledge that, the home advisor has of the user. So it's contextualizing, you know, numbers and charts in the con in, you know, in the context of that person's life. Your family's immediate well-being is a priority. Yeah. It's just comforting. It's reassuring. That make sense?
George Wang: K. And the last part that the team spent a bit of time on is, the scenario cards. Right? In the journey, we we showed three ways of being able to, play with the scenarios. So let's assume they generate these scenarios, and here's the current mortgage. Here's the three that I generated. Di erent paths. This one has cash out. It's just, you know, d v like, what I'm showing you is is just that the numbers and the the variations will change depending on who the user is, what what they need.
George Wang: But what I wanna show you is the different ways to adjust these, scenarios. So there's the sliders, and when you slide the sliders, it A ects other sliders. So it makes that trade off, very experiential. And you can pin one of these guys, like, I wanna be done by twenty forty three. I don't I don't wanna do it earlier or later. Let's just that's a good time. Then the others would change. Right? I can pin that. But while I'm pinning that, I can, you know, I can still slide that, but the thing that's been pinned, it just won't be a ected passively by others.
George Wang: Alright. So it created a new version. You can rename that version and Yeah. So it'll always just create new ones. You can hit this, and then you can, you know, create another new version. It's just like generating images. Right? Not that different. Alternatively, you can combine what, you know, things that you are looking at that that seem, palatable to you. Like, this one has interest at $6.07 5. This is $6.02 5. You can combine this interest rate.
<v George Wang>With loan term of fty years. Right now, this has thirty years. Let's see what that looks like. Oh, okay. Alright. Let's create this car from a hybrid. Okay. And then another way of doing that doing this is like, this is just an example. Right? The user will maybe they'll have some choices. They'll maybe they'll just, like, type in an open box. That's something we'll experiment with as well. But if they say Wanna lower my my monthly payments.
George Wang: And then the AI will say, tell us more. Then you say, my partner just lost their job. We need to cut $400 a month, minimum. And then explain to you just explaining to you, like, this is, this is the the scenario that we're gonna generate for you. Or, you know, you can say something else. And, yeah, it just it'll create the the cards that Your needs. And the way it's gonna be organized is that the newest one is always gonna be, up here on top.
George Wang: So you have all of these different scenarios that it that she generated, that you generated, that you can, you know, compare. And maybe that's where we get into the comparison feature of NerdWallet. You can just hit multiple of these and hit compare, and then it'll just break down everything for you. Yeah. So that's that's it. Gonna stop sharing.
Sydney Bocik: Do you wanna start?
Vinicius Ferreira: Oh, oh, George, I just want to say it was very, very nice to see your presentation. You move it forward. That's that's super nice.
George Wang: Thank you.
Vinicius Ferreira: So yeah. Yeah. I like that you start, like, going into the let's call widgets or components or something like this. Because as I was telling you yesterday, this system, we're going to have, like, conversation. And The con inside the conversation, we're gonna have, like, some widgets where people can,
George Wang: Yeah. The inline the inline items.
Vinicius Ferreira: tweak yeah.
George Wang: Right? We also need to look into generative UI. Frameworks for it, for how, like, the, you know, explanations can be generated in real time. Because right now, everything is, like, pre created, and you're just populating the the frameworks. So I see Lucas nodding.
Lucas Menegazzo: Yeah. It looks good. It's it's good to see that you are you're you are able to progress into it. It's very nice to see it. One one thing that I just wanted to point out, it's the thing I have mentioned before in chat is, one of the the things in the preview in the the original timeline, we we should have the common confidence definition, before progressing into our timeline as Bix. So we just had some changes into the timeline.
<v Lucas Menegazzo>I just want to make I just wanted to point out that this definition is very important to us, especially into, the textual side because we are, like, one week advanced from you guys.
George Wang: Alright.
Lucas Menegazzo: So we need that definition this week so we can progress into the next one with all the definition.
George Wang: Yeah.
Lucas Menegazzo: I think you progress it really well really well on the visual side. I would like to see more about how the conversation should go especially. Message length, how should be the proper, reception from the AI to the user,
George Wang: Yeah.
Lucas Menegazzo: the different type of message that we can use. So tactual side, I would like to see more of that because we'll use that, as resource materials for comparing AI. So we're basically now comparing architecture architecture definitions that we could use for AI, and we are seeing how they behave.
George Wang: Yeah.
Lucas Menegazzo: We we want to see how they Context and tones based on the common confidence definition that we are bringing in. So I just wanna make sure this is on our your radar.
George Wang: Yeah. It's processed.
Lucas Menegazzo: It's going to be very important for us in the next week, but I'm really excited to see, your progress into it. I think you're you're moving very well, into what we have.
George Wang: Yeah. Appreciate it. Any terms of the architectural decisions? I wonder if it's, like, this or this, or there's three, or it's, like, are are we talking about Parameters do we need to input into the AI for it to do, you know, understanding that sort of level that we're talking about,
Lucas Menegazzo: Yeah. I
George Wang: you know, will really help me.
Lucas Menegazzo: I think it's a bit more, how to make the abstract concept of Roam confidence into a conversation. You can do it, like, for with examples of,
George Wang: Sure.
Lucas Menegazzo: like you did. This could be, an example of conversation that would help or words, tones, things, do's and don'ts on the conversation, because this will help us to After it and to create this, specification. I don't have a, a right answer to what it is, but, ways that you can bring to us what it should look like into a conversation. The words we we could use, the how the conversation should ow,
George Wang: Yeah.
Lucas Menegazzo: this type of stuffcould help. And also the specific details. For example, how much how how should be the length of a greeting message. With WhatsApp on the past using LLM agents. And the the, ne balance between giving too much information and giving the right information is something that is very a very delicate balance, and it has a human touch to it. So how big should be the messages? How big should be the the reading messages?
Lucas Menegazzo: How much explanation should AI do? Because some of these things are going to be like Put into place. We have something that we control and something that some things that we don't. So we need to create boundaries and this type of stuff, help us. So you you cannot do, like a very long text. How much is a long text? You cannot do a a a a very big greeting. How much is a big greeting? This type of of, kind of guidelines on what is a common confidence conversation, really help us to to achieve that through AI.
George Wang: That makes a lot of sense. And the thing I'm thinking about is how do you not over specify? Right? Like, AI is getting better every week. There's new things coming out. And if you over specify something, you're you're almost,
Lucas Menegazzo: Mhmm.
George Wang: like, you see companies that are getting their eat lunch eaten by anthropic, and they just keep on doing what they're doing. They're not going back and reengineering everything, you know, and this is a common pattern. And so to to kind of Make what I'm saying tangible. It's like, if you just give it the personas and you tell AI, hey. Adapt. What are we going to adapt? Whether it's length or tone, words that you use for this persona, that is a that's that's simple, and then lets AI figure out what do I need to, you know, do what what do I need to change in order to just service this person that's like this?
George Wang: Versus I want the length of your, you know, comp I'm gonna use these kinds of words. I want you to use, this length of then that then it kind of it becomes very unwieldy. Right? And then we have to and this is something, like, we we we should experiment.
Lucas Menegazzo: Yes.
George Wang: Right? And I'm wondering if there is a and, actually, the team was asking about this as well. Is there is there an engine that we can use to trace different, for different personas? Into the, you know, the the interface and have AI be able to use the parameters that we assign it, whether it's and and, like, all kinds of permutations of these parameters. Right? Whether it's, like, more detail, less detailed. It's almost like, it's almost like, yeah, and then you have a supervisor that's that's evaluating the quality of these conversations, and maybe the supervisor is something that David and Sydney can set.
George Wang: What is it? Like, what is a good conversation regardless of the, the person coming in and their their needs? And maybe when, you know and then then you you can do you can loop these and check where is the AI falling o . What where is it? What is it doing that is not aligned with what we think it should do? And you hone in on those. That's What we're seeing as a practical solution to, you know, just do a lot of experiments and see what breaks versus over specifying everything at the front end.
Lucas Menegazzo: Yeah. We have something to compare to.
George Wang: Yeah.
Lucas Menegazzo: So what are the guidelines or what should look like? Because we are going to do a lot of testing and and and to to achieve that. So, in order to compare if the persona is being achieved, what what should it look like so we can, during testing, see if we're getting closer or away from that. We can do it, like, generically and ask AI to do so. But, I would like to have just, like, a a baseline or something to compare to so we can see if you're heading to the common confidence, type of conversation or if we are drifting apart, in some way.
Lucas Menegazzo: I know there's a lot of trial and error that we will do and just ne tune it. I just wanna have something to compare to, to see if we are, getting closer or away from that.
George Wang: Yeah. Makes total sense. Yeah. Let's continue to chat o ine about this. Yeah. It's a it's iterative process, and, you know, we'll we'll need we'll need to we will need to figure out what what what is that point where, okay, we have what we have is good enough. Now David and Sydney, here here's what we like to get your input. Now, like, play around with it. Put it pretend you're different, you know. Personas or any different things and see what breaks. But, yeah, that's that's gonna be part of the process.
George Wang: You know, we're not gonna be able to gure it out on the firsttry. Oh, Vinny.
Vinicius Ferreira: Okay. So just, a couple of, things that I want to add to your, design. Something that would be very okay. So firstof all, I like that you brought wireframes. That'sl super nice to focus on the commercial architecture. That's that's nice. Something that would be also interesting would be to provide users Some kind of, ways to get more explaining, some tool tips or something in the cards, some guy explain me.
Vinicius Ferreira: It's always good to have those kind of, things. Also regarding, the conversation, something that that David brought to us is that this tone should be kind of Super announcement people, I don't know. When they are talking to someone else, this person understanding and understand the the way, they speak, and how to respond in a way very similar or something like that.
Vinicius Ferreira: So, this is they're kind of nuances that's super nice, if you can I don't know? That's the thing that Lucas, is asking, like, some basal some some guidance for For them to, to provide the agent. Yeah.
George Wang: Great.
Vinicius Ferreira: Those are my my 2¢ here.
George Wang: Awesome. Yeah. Yeah. Those things, you know, they're always the like, yes. Yes. They they'll they'll be there. And we can only get to that through a process of, you know, back and forth and testing it. Right? Prototyping and testing it. It. That's what I'm saying.
Sydney Bocik: We do have another call after this,
George Wang: Yeah. Sydney.
Sydney Bocik: so I just wanted to give a minute or two for David to get David, is there anything you wanna add right now before we No? Okay. We have the recording, so we can, digest this a little more today. It's exciting to see the the direction starting to come out. So, thank you guys. If you guys need anything else from us today or any kind of, specific notes or comments or anything, just, let us know.
George Wang: You,
Sydney Bocik: We,
George Wang: Sydney. Appreciate that.
Sydney Bocik: can address that today.
George Wang: Yeah. And,
Sydney Bocik: Thank you.
George Wang: I guess yeah. Are we staying on or video of a question?
Vinicius Ferreira: No.
Sydney Bocik: Are you saying Mhmm.
Vinicius Ferreira: No. Sorry. I I forgot to, resolve my my okay. But, the only thing is that, for today, Sydney, I'll be revising thank you, David, for all the comments. I'll be revising everything. Okay? And I'll be, if you need it, I'll be, sending a message. Okay. Thanks.
Lucas Menegazzo: Thanks for the presentation,
George Wang: Awesome. Okay.
Lucas Menegazzo: George.
Sydney Bocik: Okay.
Lucas Menegazzo: Great work.
Sydney Bocik: Yeah.
George Wang: Yep.
Sydney Bocik: Thank you. Thank you.
Lucas Menegazzo: Bye bye.
George Wang: The recording. So alright. So, welcome to day three. So on day one, you know, we showed the journey that we did, last week and over the weekend, and we got some good feedback. And yesterday, we, you know, review the different components that we started prototyping and visualizing how the in light items, the magic moments, and how the information architecture.
George Wang: Creating these things, sort of in their individual comp parts. And then today, you know, Roam has been working a lot to see how we can take these parts and start to put it together into one system. And so that's, that's something we're really excited to show. But, yeah. Since, but, you know, before we do that The big steam, I saw you, Lucas. You you received the, the conversation guide for, calm confidence.
George Wang: Right? Did you have a chance to look into that at all?
Lucas Menegazzo: Mhmm. Not not yet, but I have this, and I'll share with the team. I brought Pedro in,
George Wang: Yeah.
Lucas Menegazzo: to this call as well because he'll be responsible for, the architecture of the LLM side. So I brought him just to have him into the conversation as well because he'll be a big role into it, but we haven't had a chance to read through all of it and to provide feedback yet. We doing this by this afternoon?
George Wang: I mean, you know, this is our definition. You know, calm confidence is like, what we're trying to capture here is the tone of personality, not the entire mechanics of how the chatbot will behave. Right? So this is just how it's gonna sound, how it's gonna feel, staying steady on our certainty, communicating plainly without dumbing things down. That's really important for us. You know, giving the user agency instead of a verdict, and reducing cognitive load and keeping the momentum. It never feels salesy. Yeah. Sure. Okay. And just for everyone here, I'm just gonna quickly go over that and be good to, you know, get Sydney your reaction to this and, you know, get a nod from David later on, and we didn't hear much from him yesterday. So, you know, curious what his, you know, what his take is on what we showed and how everything is going. But That's the ve most important bullet points that everything else ladders up into.
George Wang: There's a lot of detail, a lot of detail with interacting with different personas and what to show when. So that stuff, that is something that, you know, we can only describe so much. And at some point we just have to bring it into something we, you know, we can test, we can use that's basically, what we can show next. But, What like, just looking at this, Sydney, we included the behavior Roam, what it doesn't do, what it does do, doesn't do, the phrase bank, and evaluation rubric.
George Wang: Any any quick reactions?
Sydney Bocik: No. I like the I like what you guys showed yesterday in terms of the components, and and I think this makes sense. Like, there's nothing in here that really gives heartburn from a full, definition. I, from from the initial definition, I haven't had a chance to review this in its entirety, but initial reaction,
George Wang: Mhmm.
Sydney Bocik: I, I agree with what's, what's up at the top Roam.
George Wang: Okay. Yeah. So just take your time. Yeah. The team spent a good amount of time just, you know, really trying to craft this, as, you know, precisely as possible. But, ultimately, all of this is just describing. And until you Okay. You know, how far are we from like, how does this actually translate into how the agent behaves? So, that's why we so I'm gonna you know, we, give me just one second.
George Wang: I'm trying to nd my browser where I can walk through this prototype that we put together.
Sydney Bocik: Mhmm.
George Wang: And I see Charlie just joined. Hi, Charlie. Hey. Good. So On-It sure if you heard much about the Roam, George, with Everco, and we have Christopher as well. And we've been, we this Monday, past Monday, was the the firstday, o cial day of the project. And, we're really excited to, you know, work for the next three, four weeks to, bring AI into the mortgage industry.
Charlie Gann: Yeah. Thank you for that recap. It's it's great to meet you guys. And, Sydney, of course, as as you guys know, keeping incredible notes, I've been able to sort of follow along from behind the scenes.
George Wang: Awesome. Okay. Alright. So, I mean, there's a lot of like, I was saying, there's a lot of information descriptive, like, how the chatbot should behave. Right? So, you you know, we just the best way to know how if it if if we get a right description is just to test it. So, this is what you see. Just completely wireframe. Anything can be changed.
George Wang: So this is really a session for us to get feedback from the team. So let's say, like, hi. Let's just say hi. Right? We don't need to give an address. I I gotta be asking about anything like, you know, let's just say I heard a race dropped. Is it worth looking? Crossing my ngers that this works. Okay.
Sydney Bocik: No. I like the I like the prompts. It's helpful because I feel like a lot of people a lot of consumers and stuffin mortgage are not sure what to ask, and I feel like that's a lot of, a lot of the issues with Chattopting, being required to prompt it for something like this. So the The prompts having prompts available, I think makes a lot of sense.
George Wang: Right. There are so many things that we could be prompting the users for. Right? Oh, Lucas. ffi
Lucas Menegazzo: This is some some type of useful, insights that if you guys can document later will be very helpful, like providing some options, like, Vinny has added it to the the the, his ows, like, for example, having reply buttons. So take a look on what Vinnie has created in terms of the ows. There are some elements that he put there as suggestions.
George Wang: Yeah.
Lucas Menegazzo: This this type of structure helps us a lot when developing the architecture later.
George Wang: Yeah. Yeah, absolutely. That's, that's the whole point, right? We wanted to hand o at the end of this week with something that's more concrete and everything all coming together rather than just, you know, more descriptive things so that you guys can start with more confidence. Yeah. And like the the and Vinny, yeah. We'll we'll chat later. The prompts, I, you know, we're wondering, I mean, there are so many things that you'd be prompting the users for. Right. And, and that's, that's something that, you know, we'll, we'll log as a decision that we'll make, you know, past this.
George Wang: Like, what kinds of prompts do we wanna give the user at what stage of the journey?
Sydney Bocik: Yeah.
George Wang: So let's say I enter this address. K. I'm entering the address, and it's just pulling estimates on the home value balance, and I can see more details. But this is all pretty straightforward. This is is the Adam stuff. And yeah.
George Wang: And you know what? This is where you get a place to start with your home address. Okay. So I asked where this rates drop further. It's telling me it's guiding me through. This is all a work in progress.
Sydney Bocik: Mhmm.
George Wang: It's like the the the surface that we can use to prototype something. And then based on what we get back,
Sydney Bocik: Mhmm.
George Wang: we can iterate. It's like, okay. Do we like do we like, what we're seeing here? And, you know, can you explain, monetization? This is a magic moment. This might not be coming up the you know, it we get to decide when this comes up.
George Wang: Right? It could be taking up the whole screen. It could be taking up half the screen. But and, like, hey. Can you create a, visualization of breakeven? Okay. So in order to do that, it's, you know, it's basically asking me telling me the link to create different scenarios.
George Wang: Okay. Create. Scenarios. Create di erence. Re scenarios. Then what it's gonna do is pull up these different scenarios, and I can select any of these and change the Parameters, or I can, you know, combine some element from I mean, this what you're seeing, it's it's just an example.
George Wang: Right? Like, you're, there's gonna be more elements that you can combine. We have to think through the logic that, you know, but this is the overall architecture that I'm showing you guys. And then and then you can,
Sydney Bocik: Mhmm.
George Wang: you know, use that to create a new scenario. And then okay. Yeah. Explain amortization. Show me a visual. We'll have to teach the AI to know when to show the visual, when to, you know, show a different version of visual, different version of the visual. So all of this is getting generated in real time.
Sydney Bocik: K. Very cool.
George Wang: Yeah. So that is hopefully and then, you know, on the left side, we have this data accuracy graphic. Basically, it's like signaling to you without putting too much pressure on you for, to provide data. So the more you can and then we'll yes. Tool tips will be part of all of these things. But, know, at this point, we're really just focused on the spine of the experience and, like, what are the big components and how these pieces talk to each other.
George Wang: You got questions, Charlie?
Charlie Gann: Hey. Yeah. What at this point firstof all, super cool that you've stood this up in the short amount of time that that you've had, so it's really neat. What sort of feedback at this point in time would be really helpful for for you and and the Roam. Like, what are you kinda looking for from us?
George Wang: That's the like, more so the the, the overall direction.
Charlie Gann: Link.
George Wang: Are we going the right direction?
<v Charlie Gann>Okay. K.
George Wang: And there's a lot of details that we, of course, need to adjust, but showing the different scenarios in the middle, you know, your your, everything about you on the left side, the conversation on the right side, and generative UI, creating visualizations and the magic moments and how these pieces all t together. That is the that's the direction. That's the architecture that big Roam needs to move on to, you know, to to unblock them, from what they need to do.
Charlie Gann: Cool. Yeah. Thank you. So in this in this layout, let's say I'm going through this. This is pretty heavy stuff. You know? For me, personally, I tend to look at it and stare at it, think about it, and then I get distracted, you know, like, just by something. And then a few hours later, I wanna come back. Is there where's the point of reference? Is there, like, multiple conversations that'll happen that we can go back and reference, or will will always return to the context that I left?
George Wang: This is where, based just in the journey map that we showed earlier, once you get to a point where, you know, you can you can you have the option to save your progress,
Charlie Gann: K.
George Wang: You know, where where like, that feature is there, but, you know, where is the point where we invite users to save the progress? You know, is it after the scenario or scenarios are generated? Is it after the the Roam? I think, you know Personally, I think of the scenarios, but I gotta talk to, you know, our team and get but, you know, kinda curious. Does that make sense, after the scenario is generated? Or Mhmm.
Charlie Gann: Yeah. Great question. Yeah. So sorry that I missed the context. So at this point, has someone already created an account?
Sydney Bocik: We that is David,
David Kawata: Caroline, have you read the brief?
Charlie Gann: The I'm I've read a couple of briefs yet.
David Kawata: Well, so the the brief in the beginning is the account creation is actually kind of buried in the process. Right? So Remember I think,
Charlie Gann: Right.
<v David Kawata>George, one of the things that I I like to have to be thought about is that somewhere in the visualization, whether we ask for initially, there would be an indication of, say, your save your progress. Right? Because we have Three columns here.
Charlie Gann: Mhmm.
David Kawata: So to imply that if they left that their progress wouldn't be saved, and so always inviting them to have the ability to save or and then from there, the prompt to create an account. But then that would be my ask for you to noodle with your team how to do that in an imitation way and also warn them that all this work that they're doing, potentially, before the direct ask, that it could be lost. Right? So, I think that the balance of, seeing the sign on the door for that that they wanna get to, but not, like, be aggressive.
David Kawata: So thank you.
George Wang: Right. Right. Pedro?
Pedro Vieira: Awesome. So firstof all, really neat demo. Really liked seeing it. It's my firsttime seeing it. I think I saw some statics, yesterday. I think that the main, feedback slash discussion here is I thought we're going a direction where we're going to be mobile first. And then then we spray, desktop horizontal landscape, heavy. Right? So it will be really useful if you could see. I'm not expecting for this to be responsive right now. We explored firstin the mobile, and then we can maybe expand in the desktop because the navigation between those scenarios, how we're gonna, you know, change between those screens, that that's pretty important for us.
Pedro Vieira: So just this directional adjustment that I I'd like to to chime in.
George Wang: Yeah. That's good. I mean, the team has been thinking through that too. I don't have it ready right now, but it's definitely something that, yeah, we'll get aligned on that. David?
David Kawata: Again, the the low delity, again, so, George, I I love the the toughness, the the interaction. So these are those are great. I I do have this I'm gonna assume the design itself is not the firstfocus. Right? It's the the spine, the intellectual, the interaction. Right? Right. Perfect. I I just wanted to highlight that what I see here. One of the things I I really I really believe me. You'll tell me if my if my feelings and my little baby
George Wang: Right.
David Kawata: research is right. But aesthetically, having a a feminine I mean, I wanted to say it. I don't know how politically correct it is. But I I believe that it needs to be pretty. And for, like, a 30 year old woman that lives on her phone aesthetically in the design because I think it it just sees Again, that that was my firstkinda look on the look. Right? But I know it's low delity. So I just wanna reiterate that the animation, the evolution, the graphs, the sharpness of this seems like there's, you know, it is very functional.
David Kawata: But I wanna make sure that we
George Wang: Oh,
David Kawata: the spirit is inviting and and ready and disarming any oh,
George Wang: absolutely.
David Kawata: okay. Cool. Go ahead.
George Wang: Yeah.
David Kawata: So Perfect.
George Wang: That's part of the week one deliverables. Over the next two days, that's where we, you know, really Just black and white and no design, no corners, just to, you know, not for us to get distracted by all the the the key these two conversations separate.
David Kawata: Perfect.
George Wang: So that's something we're working on. Yeah.
David Kawata: Thank you.
George Wang: Yep. Vignesh. Vineet.
Vinicius Ferreira: Hey. So, my my then sign it's looking here. So okay. So, right now, we have a lot of information. It's a kinda overwhelming. So it'll be very nice to, to to have everything in in the, chat timeline. And, also, I'm concerned about, consentments. Is it something that we need to, also pay attention,
David Kawata: When you when you say consent?
Vinicius Ferreira: as for the yesterday as well?
David Kawata: You talk about consent?
Vinicius Ferreira: Yeah. Yeah. But make making
David Kawata: Okay.
Vinicius Ferreira: that transparent for users. So we are gonna, look out for data. I know I know. We're gonna, try to nd data. Something like this or, and, also, George, did you check okay.
George Wang: I see I see Pedro raised his hand.
Vinicius Ferreira: Did you just one more Roam thing. Did you check the CFO website?
George Wang: Yes.
Vinicius Ferreira: Yeah. They they have a lot of, visualization cards over there, like, some examples. This is maybe that could be nice
George Wang: That would yeah.
Vinicius Ferreira: to check.
George Wang: Which CFO website are you referring to?
Vinicius Ferreira: It's a website. It's called cfox.ai.
George Wang: Mhmm.
Vinicius Ferreira: It's on, it's on, the design kicko . It's one of the I think this is the last one in the the competitor. So that's something.
George Wang: Oh, okay.
Vinicius Ferreira: Yeah. They they have a lot of examples.
George Wang: Yeah.
Vinicius Ferreira: It's mostly focused on financial advice, advertise sorry. It's a an I financial adviser, but it's it just, show, some some cases that are very interesting for us as well for us to understand.
George Wang: Examples that we can, incorporate into what kinds of visualizations that we provide, our users. Right?
Vinicius Ferreira: Yeah.
George Wang: Yeah. Let me check that about chat about that.
Vinicius Ferreira: I think that having,
George Wang: Yeah.
Vinicius Ferreira: scenarios, having, this data, I think we we can have places we can go there and check, but I don't think that should be, like, every like, shown all the time. It's just, we can have, like, places where people can go and check or something like this. And also about the, The the moment. I don't know. I think that, just showing people, like, just, is I don't know.
Vinicius Ferreira: It it can be also something not good. Like, maybe inside of the conversation have something, with some kind of animation there. This can provide a lot more natural, feeling or some good feeling.
George Wang: Yeah. Absolutely. Yeah. The example that, you know, we show is just okay. Here's the spectrum from just full black screen in your face. Of course, you know, there's the visuals inside. There's the black boxes. There's all these things that we can do. And Yeah. This is just the firstweek, and we just wanted to get the architecture right. Like, we heard you match moments, and now it's about re ning that, getting to, you know, how do we show that without being so much in your face? And what what should be, like, really in your face, you know, and tweaking that dial is, is, you know, the job that we need to do.
Pedro Vieira: I'm just gonna space yeah.
George Wang: Pedro?
Pedro Vieira: Just gonna expand a bit on the consent commentary that Vinnie made. The thing is that we're gonna pull credit from people. We're gonna, reach to third parties to get more information on them. And some of those ows might require some consentment to some consent to be given by the user. Right? How do we prompt that? I think it's a it's an art in itself because it's a Sometimes to be, as the additional signature or something like that.
Pedro Vieira: So where we wanna put that in the ow, how early, how late we do it, that's, I think, a product decision that we need to make alongside the e compliance decision, but also a UX decision. Right? How we we do not we expose the the user to that kind of a friction without, like, making them less, happier or excited to what's coming.
George Wang: Ja, der.
David Kawata: Yeah. So on the the consent credit, other things, I think that would be in the terms of services on the account curation. Right? You you can't pass go. It looks like $200. So you On-It be able to run Less there's an account creation. That's the saving parts of the process and the MFA process. So I think, again, to your point, what is the call to action? When is that embedded? How is that experience that right. That's some of the magic you guys will gure it out. But then this end of consent, explicitly permissions, and then that would be in the terms of service out of the account creation, at least, so just my 2¢ initially on the terms of service.
David Kawata: But then In the different veri cation services itself may have an additional opt in process, and those each of those digital doors for those services probably has their own prompted, consent. Because again, to your point is the friction point. So we don't wanna ask three consent con rmations when the consumer is only gonna do one. But the terms of service probably at the account creation should have our general, disclosures. That also reminds me that you probably need to get some example terms of services.
David Kawata: So we'll work on that this week. Okay?
George Wang: That's good. Okay. So I'm really glad to get everybody's feedback and alignment on this vision. There's a lot of, you know, me. We gotta put on the bones, but I'm glad that we have a direction that we're, you know, marching towards. So the priorities for the rest of the week is the visuals, and Showing what the responsive view will look like, so the VIX team can, you know, just there there's there's, there's no ambiguity on that.
George Wang: Any other questions?
Sydney Bocik: did you have another question? Or I had a question. Okay. I didn't know if your because your hand was still raised. The the question that I have and this looks great. Obviously, there's a lot of property data and and data about the, for the scenario cards as well too. Seeing this visually will help us kinda go through now that we've seen an initial layout. This will help us go through and, from our perspective, help refine the data that we think might really be important at these stages.
Sydney Bocik: I intentionally tried to stay a little bit out of that in the beginning because I wanted to see the starting point. But how, what I didn't wanna do is say I think all of this data is Overwhelm the user. So is there should we start thinking about the stages that exist in the user journey maps that, like, Vinny, that you had worked on and start pulling together contextual data around what should be shown at each of those stages, what's important, or what how can we how can David and I be helpful?
Sydney Bocik: Around, what inform like, contextual data and data from APIs, data that we might know about the user at each stage. Is that necessary now? You guys mhmm.
George Wang: For the firststage, it's correct me if I'm wrong, but there's just Roam data. Right? Past the stage, there's veri cation. There's other APIs that we we we we, integrate with. But Like, I think it's just a it's just a matter of having a have a conversation about that. Like, what do you think? But, you know, just for the purposes of this meeting that's, you know, that's just ended here, and we could, you know, set up a, you know, deep dive on, you know, like, what are the different states, that the user might be in, and, what kinds of data are helpful, when they're in those states.
George Wang: And there there's a lot of different pieces that you gotta you know, we need to try and emulate. Right? Like, this user journey, you know, it's So far, it's like you gotta balance things out on a temporal perspective. You gotta you know, there's all these different dimensions that you gotta be looking at the same time, right, to ensure that it's a great experience. So we'll have we'll talk about that. And, David, you had a question?
David Kawata: Yeah. I was just I think extending I don't know what was in the to the third party data set, in that, but as regarding information, options. Right? The options can be overwhelming at a point unless asked for comparison. So, again, I I think you understand that. I think we're we're moving directionally ne. I think, you're you're getting the components that are going to be available when they are appropriate, and then we can start pruning and creating more restraint.
David Kawata: But I think, Sydney, you're really more talking about that overwhelming of options, overwhelming with details, overwhelming until that kind of meet meets the validation of the consumer wants that. Is that is that really what you're talking about as far as data?
Sydney Bocik: Yeah. And I is yes. Because I also think that if David and I looked through the data through the APIs coming from a mortgage background, would say a lot of the data is important and to the user. But I know that what we might think would be important to pull in from a mortgage background would be really overwhelming. So I I I want to make sure we're getting you guys what we you guys need, but I also wanna make sure that we're not pulling together so much that it does, that Roam a mortgage background overwhelm the user in terms of what would actually be important to the consumer, not not like a mortgage professional.
George Wang: Yeah. And this is where I'm glad we have AI. Right? There's a lot of detail that AI is able to capture. And, you know, if their user you know, we might need different scenarios for you guys to specify if somebody's in this scenario. Like, this is what I would show them. Just like how we did the, you know, the conversation guide, the reference, more data we have, the the more we can instruct the AI to, you know. Do this when hey, David.
ff ff <v David Kawata>One one thing I I it's a I perfect. I On-It wanna just ask again for reference of in the back of your mind. I'm super conscious or thoughtful in the sense that I don't wanna build a feature what OpenAI or or cloud could use. Right? And so I want George as again, we're looking at competitive landscaping from your consulting background. Say, alright. Are we building a feature that next week, Claude could just Open up and kill us in the sense of value to value.
David Kawata: So one of the thoughts is as we work through, not this week as the deliverable, but I I actually would love to get your thoughts of, like I have some opinions how we are different than a a public LLM service, and data hydration and emotional layers are important to that. But, again, I'm gonna just ask for the team to noodle. How do we protect ourselves from being a feature upgrade from these, That we can really differentiate as value. Like you said, AI is amazing, but we're not the trillion dollar company.
David Kawata: We're using the trillion dollar company. So that that's, one of the other interesting opportunities and challenges Yep. That I that I, present back to the team.
George Wang: Yeah. And the initial thinking on that, I will add more, details to this, but the initial thinking is, just this, what we see right now presented to consumer one on one. We already seen so many companies take certain industries, use build things on top of the existing LLMs, and the LLMs would just, you know, make them obsolete. But the thinking behind Connecting with the consumers and connecting with the, the CRMs, and connecting all the dots in this industry, creating that ecosystem, that is what's ultimately gonna create the defensibility because it's it's easy for Chargebee to to create something to as, you know, tying to their tying to Adam and answer somebody's question, but it's harder for them to You know, if that person already if that person needs to submit something to the loan originator and the loan originator is working with the broker and there's all these peep people that are working together using the same platform, then it's gonna be, it's much more defensible than just translating, API data.
David Kawata: Yeah. No. I I I, again, I agree the the the connectivity and really the actual real precision of the data. Right? So if you go to chat g p t and you look up a property address, their access is really Zillow, and not necessarily updated current information on images and or public or or more data. So, right now, there's a there's an awareness. I just wanted us to kinda chat about that and always be paranoid about That we create that is aware that the consumer goes, okay.
David Kawata: This is different. This is worth me giving more information that is safe. So, all that consideration. So thank you.
George Wang: Yeah. Absolutely. We're not building a feature here. Like, we wanna prepare to build a company that's defensible. So, yeah. And, we'll we'll coordinate on maybe we can set up more time to to chat about that. Right. Thanks, everybody. See you tomorrow.
David Kawata: Get that automatic yet.
Sydney Bocik: No. I keep I gotta look in the setting today.
George Wang: Guys, good morning.
David Kawata: Morning.
Sydney Bocik: Morning.
David Kawata: How are you doing,
George Wang: Back from the conference.
David Kawata: George? Yes, sir.
George Wang: How was it?
David Kawata: Productive. Yeah. But still Vegas.
George Wang: Nice.
David Kawata: Still Vegas. Still conference.
George Wang: Yeah.
David Kawata: Yeah.
George Wang: Nice.
Lucas Menegazzo: Guys. You'refine.
George Wang: Hey, Lucas. Lucas, did you have chance to review the conversation guide from yesterday?
<v Lucas Menegazzo>Yeah. So I'm putting, one document, together with the team, just with a few open questions, let's say, about things that I would like your input into.
George Wang: Yeah.
Lucas Menegazzo: They're more like Were, some, like, dead ends that we might face, and I would really want your opinion on it. We'll be nishing up by today, and I'll send her over later.
George Wang: K. Sounds good. Alright. Yeah. So, I mean, did you get a chance to do you guys get a chance to play with a prototype?
Lucas Menegazzo: Yeah. My my point here is, I think Pedro mentioned this. Yesterday. But, on regarding to have, this, in mobile, I know you you guys have, like, three three panels, and the conversation sometimes go to the the third one, and their information all over the screen. I'm just curious to see how you would do that in mobile where you don't have,
George Wang: In mobile?
Lucas Menegazzo: yeah, that much of, for horizontal space. I think that that's my main feedback into it.
George Wang: Yeah. So that's something that we'll we'll show today.
Lucas Menegazzo: Awesome.
George Wang: We've experimented with a couple different ways to, squeeze all of that into a mobile interface in a way that it still works on the browser because we're not you know, eventually, maybe we'll build in that. But for now, we just you know, we wanna be the user to be able to navigate across these three sort of Panels easily. Yeah. So we'll go through that. And, yeah. You know, this is day four, past couple of days we created the journey, put the experiments, put the, components up on Apollo line.
George Wang: Everybody saw that. We recorded the prototype and, we're working. The team is working on Till tomorrow Roam show that o .
George Wang: And, so in terms of mobile view, let me just share my screen, show you what where we landed on. Okay. Here. So yes. See. Right? This is the this is recognizable. This is the mobile. This is the mobile view. When you collapse the page to a certain, you know, width, it just collapses into the the three and a half three and a half bars.
George Wang: Three and a half and a half bars. Yep. And, actually, let me just Putting together the visuals for, for, like, a visual direction. And we heard from David yesterday that, you know, it's like a effeminate 30 year old, women. That's the target sort of aesthetic. And so we're aiming for that. We're gonna show that tomorrow. The team is, has made progress, but, you know, you have guys haven't worked with Christopher before, but, you know, this man has pretty high standards. So, we're gonna See what that process looks like. So right now, when you start the site on mobile, you're in the advisor view and okay. K. And as soon as, you know, your address is found, this becomes active so you can just you know, you can go to home.
George Wang: Maybe we'll Add some visualization to make it really obvious that this panel is now active, and then it just, you know, it's almost like a, something that hints that you can discover it as you have conversations with the advisor. And you just say show me scenarios, and then this should become active as well.
George Wang: Makes sense to everybody?
Lucas Menegazzo: Yes. That, thinking, kind of speech bubble is something that we were also after. I'll I'll put it as an a open question. I don't need an answer right now, but, one concern is some some parts of the that more technical part will will take some time to load. Like, for example, the images are not instantaneous. Over to the user.
Lucas Menegazzo: Like, how should be the the expected behavior or what do you think will be a good behavior for the AI, to just pause, input for some time? Like, when you use cloud or use chat GPT, it get it it stops the input or adds to a a a queue? Or, should it allow users to keep typing and it keep answering and later when he has the the Artifact generated,
George Wang: You mean this?
Lucas Menegazzo: each presents the user.
George Wang: Where you like this? The image?
Lucas Menegazzo: Yeah. Like this, for example. Yeah.
George Wang: Yeah. Like, this is all live. So I'm this is all connected to Adam already. So it's pretty fast usually.
Lucas Menegazzo: Mhmm.
George Wang: But, you know, in case it hangs, just, like, you know, show the, like, the bars where the text would go or where the image would go. And then when it loads, then you put the image in there. But you just wanna show people that, hey. Something's happening. It's loading. Like, when you yeah.
<v Lucas Menegazzo>Mhmm.
George Wang: You know, Yeah. So that's that's that's, you know, the the the the tactic for components that get shown. But, you know, we all have experience talking to an agent and, you know, we we like it's it's expected that it'll take some time, to get back. So it's just when the components get shown, right? That's what you're referring to. What's the mhmm.
Lucas Menegazzo: Yeah. I I just wanted your take into it if it's that that that's the best approach or if you would suggest something else. I just wanna hear from you what what are your thoughts on it.
George Wang: Yeah. David?
David Kawata: Yeah. So that that's I think it's a really great concept to to just address. Right? So if there's any known delays down the stream of collecting data, the calculation of the formula that we know in our interaction that there's a a text communication that things are loaded in the back where the customer's not waiting. Again, not necessarily. I think that's what I'm hearing, Lucas. Right? There may be a lag on the population of the magical moment, and the wait will dilute that customer experience if we know that there is a potential Your point, George, you're testing this out.
David Kawata: The all the Wi Fi works. Right? It's cool. And but in the native, there may be a moment that we understand. It may be a little bit of connection issue pulling information. So just if we could be mindful with the team, at their certain moments of data collection, that we may engage the customer with another, hey. That's great. Thank you so much for sharing, or some type of positive interaction where the customer doesn't know that we're doing something on the backside, that magic. So I think that's a good point, Lucas.
David Kawata: So as we move down the stream, is there a recognition to the user experience that shows progress and but yet our system is working on behalf for them, and they don't understand how the sausage is making or the how it's being cooked. So cool. Thanks.
George Wang: Yeah. Any other questions regarding responsive? Okay.
David Kawata: I I would I would, again, it's not a need, but on the animation and the transition, if there was anything thematically into a home neighborhood that process.
George Wang: Absolutely.
David Kawata: Right?
George Wang: Right.
David Kawata: So instead of a ball or dots, if there was some other ways, we could put something charming there, that'd be cool. Bye.
George Wang: Yeah. Yeah. We want everything to be animated and alive when we ship this four weeks from now. So on this point of, you know, we created this reference for the engine for VIX to, you know, The agents against and tested, and there's this cycle that needs to happen. It's right. Right. It's like what kinds of prompts do we put into the agents, and what is the behavior that that's gonna drive?
George Wang: David, you raise your hand.
David Kawata: I'm sorry. So on a conceptual idea for young people like we have on this call, that, like, show me scenario in the text and help me to understand for ADA, for older folks, for right?
George Wang: Yeah.
David Kawata: How would again, I know that this is low delity, but, is there is there an approach that the user would be able to if we do hyper hyper hyper personalized and we know that certain age, this font size is really cool, but To read. Help help me understand how if this is the approach that you're looking at, or is there some type of interaction or transition, for certain demographics? Or if it comes customer wants to opt in to a certain style, that would be perhaps bigger fonts, bigger buttons.
David Kawata: How how many know how we address that issue?
George Wang: Mhmm. Yeah. We're designing the everything, against accessibility guidelines,
David Kawata: Okay.
George Wang: best practices. That said, you know, older, folks do want like, my mom, she she has she can't read. She wouldn't be able to read regular everything is magni ed.
David Kawata: Right. Yeah.
George Wang: Right? So, now we're we're you know, that's part of the accessibility. Workstream, we're always, you know, exploring, like, making the fonts bigger. And some browsers allow you to do that. You know, has those features to to enlarge the fonts just by default, by the the settings of the phone.
<v David Kawata>Yeah. Perfect. I just wanted to clear that up where you
George Wang: Yeah. Yeah.
David Kawata: I I will tell you when right. Everyone in this on this call in the write up generally conceptually thinks about a younger person journey, not wanting to talk to a loan officer that that right. However, one of the use cases that I wanted to be clear is actually actually when we are successful, create a lending as a service and a loan officer, my goal is to build a personal CFO, and that will be in more wealth management. So that'd be a top down approach. Experience to try to make technology a little more, accessible for a certain generation.
David Kawata: So both sides of the of the age spectrum. So thank you.
George Wang: Yeah. Yeah. It makes sense. Vinny?
Vinicius Ferreira: Actually, I have a question for David. David, since we are talking about, since this, UI is conversional UI, is this, AI agent going to have a name?
David Kawata: Yeah.
Vinicius Ferreira: Or mhmm.
David Kawata: Well, yeah, we'll we'll definitely look at naming and personalizing it. And, right, I think that is the the the best practices today. Yes. So, like, we we've, had, Rihanna. We've had other different we've had Rex. Again, we could go into the naming convention, but yes.
Vinicius Ferreira: Okay. So the next question will be, will be, multiple agents. Something I was thinking also something related to what Luca was, asking is that about, imagine that we need to pull out information from the system that gonna take a while to do that. Was thinking maybe we could, during this process of waiting, the agent can say, oh, I'm gonna ask for, John from department x to, check this information and then suddenly and then the conversation continues and then suddenly, oh, John, has, like, the response or got information that we need.
David Kawata: I think the storytelling of how people and the team and the and
Vinicius Ferreira: Okay.
David Kawata: the services working for the consumer, that might be a fun story, level up. But I I don't know if that's distracting them from this point of where we're at. But, yeah, I again, I think the story that we have a team working for you, for the Wang family, for missus Wang. ffii Right? What whatever it is, I think that is a, rate. We talk about social proof and influence and and being able to serve and all that concept. Vinny, I think that could be a fun story. But
Vinicius Ferreira: Okay. My my last question. Okay? Okay. We we have this system that, provides value to the user without going to, BII. But, I don't know. Are we going to have a kind of, like, handoffto a human? Something like this. Imagine that the system advised, or you need you need to,
David Kawata: The short answer is yes.
Vinicius Ferreira: discuss that.
David Kawata: So the short answer is we will need to have the ability to go man in the middle, right, or human.
Vinicius Ferreira: Oh,
David Kawata: So the short answer is yes. So when I use the concept or the the naming convention, digital lending and highway, One of the reasons why I like that analogy, you have multiple different on ramps for the customer and their journey. So let's say, we got George, and then we got George's mom. Right? They have perhaps two different pathways for homeownership. Right? But the the Fastlane and autonomous Complete self serve. The middle lane, maybe a hybrid, man in the middle. And then the slow lane, fully white glove human service.
David Kawata: So that's how I'm visualizing the functionality of that analogy. Right? So we have multiple on ramps with the consumer journey. They can and they need to get be able to get o , then come back on, and we need to know them. Right? We need to have the autonomous hybrid white glove, and then we need to have all of the to your last, your firstquestion. Right? So if you have John reaching out for property information or tax information or insurance, right, that concept of multiple on ramps, but the customers doesn't get interrupted.
David Kawata: So, yes, man in the middle hybrid process will be necessary moving further.
Vinicius Ferreira: Okay.
George Wang: Yeah.
Vinicius Ferreira: Thanks.
George Wang: And, you know, just given the diversity of personas that we wanna serve and, you know, different on Roam and off ramps and the just we really wanna, see where, our design for how the agents, behaves as in the prompt and, like, all that prompt engineering, like, how that, would stand up into all of these. So so that's why we thought it would be actually, let me just keep sharing that screen.
George Wang: We'd be it'd be helpful to use some sort of, agent evaluation sandbox. So, the team had stood this up, yesterday. And this is kind of getting back to is, Point, I was making yes, yesterday or the day before about how do we get the how do we get David and Sydney, collaborating in this process of, you know, like, like, shaping how the agent behaves, and and there's this feedback loop through all of us and, you know, create this allow us to all align with, you know, like Like, the the desired behavior and then the actual behavior and how to close that gap.
George Wang: Alright. So this is the this is a prototype that we were putting together. It still requires a little bit more re ning. But, basically, on the right side is a copilot that you can ask to use this tool. So the you can de ne The prompt that was loaded in and all of this is just for, you know, just for testing.
George Wang: Right? For to to show you guys how this could work, exactly what Roam you put in there that that's gonna be constantly changing. And and, in a PRD, there's architecture as we have an Oracle agent. We have a couple different agents that's working together. So this is just a very simpli ed, you know, use case. And when we have these multiple Agents, working together, then, it's something that the tool should be able to be, you know like yeah.
George Wang: Hey, Lucas.
Lucas Menegazzo: George, just to give you a directional, this is something that we are putting together already, and we are a bit,
George Wang: Yeah.
Lucas Menegazzo: far further away from from you guys, a bit more advanced into it. What I really wanted you to focus is to, like, refine this, this definition and go a bit deeper into it. I know there's a lot of testing that we'll be doing by next week, and we we already have the architecture to do so. So I wanted you you to really focus on the More conceptual or more, word let's say wording of this, type of of, of definition, more textual definition because we'll be doing this kind of testing, ne tuning, using multiple agents in multiple models, in the next week.
Lucas Menegazzo: So just so we don't, overlap into The the the things that we are we are creating, I would really want your focus to be more towards, how can you make sense, to us humans on what should be the concept so we can properly evaluate the AI based on that. So this is more of what we are expecting as a deliverable for you from you guys so we can, check.
<v Lucas Menegazzo>Having having some reference on on what should be a common confidence chat, and we can evaluate through that. So just a directional on what we are expecting.
George Wang: Yeah. Hey. No. Really glad that you guys are working on that. I think what's important here is for, you know, all of us to be on the same page. We're like, On-It just hand o , like, the here's, here's the direction, and then, you know, you guys are doing the agents, and then, like, we wanna make that feedback loop as quickly as as tight as possible. Right? So if there's any way that, as you're doing the testing, using the the the setup tools that you have. You know, how does the the rest of the team or how does Sydney, get the results of those tests and be able to play around with it.
George Wang: Right? Like, because otherwise, you know, this waterfall,
Lucas Menegazzo: Yeah.
George Wang: waterfall process of de ning the direction, testing, it's like, where does it come back?
Lucas Menegazzo: We we do that. We are we're we're, we already have a front end and it's assessable. I I just really need, some directional in terms of the interaction that is expected from the AI, the guidelines that, are expected from it, the tone, how it changes, how it, interacts with the user so we can plug it in with, what we are currently developing. So this is what, I'm I'm reinforcing with you. In terms of this this was expected before, we begin our engagement, and I know we had some timeline changes.
Lucas Menegazzo: So I still have questions and doubts on what is a common confidence, conversation, what are the expectations that we are setting up, or how it should feel like, how it should look like, so we can properly test. Test is very important. We need to do that. I agree with you. But, I don't really, have a clear view of what a common confidence chat should look like. So we can sync up about this.
Lucas Menegazzo: We can chat about this, going deep, but I I still have questions. It's it's kind of, a bit shallow for now, and I would like to have more, like, information to work with.
George Wang: Hey,
David Kawata: Cool.
George Wang: David.
David Kawata: So, no. I I think that's good. Again, I appreciate the conversations, making sure that we could from both teams. With that being said, does, Lucas, have you shared the access to George? So because on the cycle communication, whether he's starting, which he didn't know that you started. He started because he knew that it was needed, but you started, and now you're further advanced. But have you shared George and his team the access so they could get that loop?
Lucas Menegazzo: No. Not yet. I was waiting on the definitions to do so, but we can share later.
David Kawata: Okay. K. Please do that. Thank you.
George Wang: And and this loop, it also should include David and Sydney. Right? Because you guys have way more expertise,
David Kawata: Yes.
George Wang: twenty something years in the industry. That was really where in your judgment is is, you know, better than our judgment in terms of, like, how should the agent behaves. And, in the Notion, I saw some transcripts, that UNC needed. And then he was showing me a couple And, you know, figure out how that can, contribute to the agent design.
George Wang: But just to kinda show you the rest of this, I think there's a lot of experimentation that, you know, we should do with the different, personas and the ways to segment these personas. From the communication mode, this fast lane, middle lane, slow lane to what to have a life event. And this is how we get at all those edge cases. Right? Intent depth, archetype, primary path, or there's more motivation.
George Wang: And then the personas is basically, different permutations of these, dimensions.
David Kawata: Quick. I'll just interject for one second. So, Lucas, we're gonna have, two other, subject matter experts that will be available to help with the testing loops in this process. And so we'll send, Kelly Madison's information and Ken Bart's. So we'll have those eventually to you. So, Sid, if we could, get those set up, per line or emails. And then when we have that environment access, we'll get them next week too. Be available for testing, please.
David Kawata: Thank you. Alright, George, please. This is a looks good,
George Wang: Yeah.
David Kawata: George, but thank you.
<v George Wang>We have a judge, and, this is just, you know, a run that I did before and just to test test test the framework. This one had nine different, personas in there, and, it highlights which one is passed, which one is failed, and then the transcript of between the Term change optimizer and the home advisor. And just so based on this, you're able to see how the agent with us with one prompt is reacting to these different, personas to identify those edge cases.
George Wang: And I've and we think that this is the way to design the agent because no matter how much you describe what an agent should do, like, seeing is believing. Right? Like, you you would rather have it just act out the descriptions. You can correct it because there's no amount of describing that will, you know, like, make, just guide the agent. You just need to see it in action and then and then, you know, provide the feedback.
George Wang: And and, and so that's that's what we're saying with the fast feedback loop. Yeah. Just Roam yeah. I'm just repeating myself now, but you get the point. Okay. So, Lucas, let's, settle some settle some time, and we can go through the system that you guys have. And
David Kawata: Where where is this? Is this where where does this experience live right now? So if Roam and Sid want to
George Wang: it's it's on local. It's on our local, system right now. We haven't been pushed it. But Yeah. You know, if it's helpful, we can we can push it up and, you guys, test it out.
David Kawata: That'd be great. Thank you.
George Wang: Okay. Yeah. Any, anything else? So tomorrow, we're we're expecting to show the visual interaction. The team is working on it. And that'll also be our alignment session. So so far this week, we have we have gone I'll just see if I can pull up the, the Roam map just to get a, get a preliminary con rmation that we're are we are on the right track.
George Wang: So we are here. We de ne the customer journey, mapped the information architecture, core ows. You have the prototype. We don't have the visual direction. We're gonna show that tomorrow. Experience principles and tone is the agent's guide that we delivered yesterday, and we're going to, provide more details, once we get, get your follow-up questions, Lucas. Charlie?
Charlie Gann: Sorry to move backwards a little bit. Just wanted to kinda throw something out there as as far as I'm just trying to process the the discussion and the way the chat's laid out. And I'm trying to, like, figure out how do we marry that to this idea of, you know, our our product really is clarity. Right? Like, that's what we've said. It's, and and as I'm having this conversation, which can be sometimes abstract, sometimes gets really concrete. How do I know as a user, like, that I'm getting closer to certainty?
Charlie Gann: Like, one of the things that I think has been so effective about a FICO score or or your, your credit score is this it's it's materializing this idea of, do I have what I need? Do I have what it takes to take this really big next important step in my life? Go get a loan for a ff home or a car or whatever it is. Your team of, like, processing. Like, is there a way I don't know if it's, you know, if there's some sort of part of the persona where we're showing, like, oh, you've given us this this information, this information, this information.
Charlie Gann: You know, you're you're, like, 70% to a point where we can con dently give you the best options possible. There's there's something about that that journey or that path that would might be really nice to be visible, as a user as I'm having this conversation. Just something I've been kind of processing of how do we materialize that.
George Wang: Yeah. Yeah. David, you raised your hand.
David Kawata: No. I I agree with that in the in the sense of the gamification of eluding that there is a gap of information. If we only have this additional piece, we can get you so much further with clarity, confidence, or what they may want. So if we use,
George Wang: Yeah.
David Kawata: a ghost scene of a house or some type of full picture puzzle pieces and we labeled it income, evaluation, property information, whatever else as you start sharing information like the address, that then it gets Your house, your financial awareness. So any type of gamification to allude that the more information is actually giving them a better picture and it's a known game. That I think that's a that that's something I've haven't communicated before,
George Wang: Mhmm.
David Kawata: but I I thought about, especially as we look at a call to action to do something like saving your progress or giving personal information that's veri ed. That's our differentiator from a chat g b t. Right? It's actually them opting into sub So if your team can noodle on that, where there is a a bigger picture that we know as an expert, we could give them everything that a customer wants. And the more that they kinda fall into the kinda treasure hunt, you talked about Costco, that that that chicken concept.
David Kawata: Right? If we could give them treasure hunt of information, now they they go, oh, I wanna nd out how to give that last piece. So that'd be great. Thank you, Charlie, for bringing that up. Thank you.
George Wang: Mhmm. Yeah. And, yeah, your point is well taken, David and Charlie. You mentioned the gamification very early on. What are those stage gates? What are those pieces? Right?
David Kawata: Ahead and try to try to give you a little bit more, of what the puzzle pieces are and and how that will give you the, pick your adventure. Right?
George Wang: Yeah.
David Kawata: Address gives us a lot of context of the starting points and and where we could cheat code and evaluations. So let me try to see if I could give you some scripts on that and then see if I could go from a stage one, stage two, stage three, and see where the personal information may provide more clarity. I'll work on that.
George Wang: Yeah.
David Kawata: You to Monday. But, I'll put that as my homework. So thank you.
George Wang: Okay. Awesome. Great. I think, I think that's everything. We're out of time.
David Kawata: Where where is this, visual, time frame this this diagram that you're referring to? Where is that?
George Wang: Yeah.
Sydney Bocik: One's in the Figma.
David Kawata: Thank you.
Sydney Bocik: Do you like to send your link up?
David Kawata: Bye.
George Wang: Oh, yeah. Oh, Roam I not sharing that? Okay. Yeah. And right now, just one more thing. Right now, we We have this data accuracy. You know, it's just our firsttab at, you know, alluding to that. There's more. But if there is a de nitive list, we can almost, you know, visualize it like a skill tree, like, in a game.
David Kawata: My my thought is is that that there might be four categories of information from a consumer that we could get that that celebration of,
George Wang: Yeah.
David Kawata: right, of completeness in different stages. So I'll let me see if we can noodle on that, and I think that would be a fun, ask.
George Wang: Yeah. Yeah. Okay.
Sydney Bocik: It should align we can contextualize that a little more. It should align as well too with the the transcripts that David and I went through where he acted
David Kawata: Okay.
George Wang: Okay.
Sydney Bocik: at ClarityEngine. A lot is in there. If you look at the way because I
George Wang: Right.
Sydney Bocik: I broke down the conversations, but the raw transcripts are in there too, so you have everything you need.
George Wang: Yep. Yep.
Sydney Bocik: If you look at the way he asks questions, in I would look at the refinance one first, but we went through refinance and purchase. It was very long conversation, so, obviously, that will need to get skimmed back in some way. But the way he asks questions, in terms of it's, that will give a lot of context into what information he's looking for, how things are asked, when it's asked, and and what it kind of leads to.
Sydney Bocik: It was an o the cu conversation, so it could have been it was not,
David Kawata: Wasn't that structured for yeah.
Sydney Bocik: perfectly script it.
David Kawata: Wasn't wasn't structured scripted to give you the tacticals quicker, so apologize for that. But, I know we're at past time. But, yeah, if you could do that, we'll try to give you some documents as a starting point again, by Monday. And I think I have some ways to hopefully, provide you some clarity. Unintended.
George Wang: Sounds good.
David Kawata: Alright.
George Wang: It's clarity.
David Kawata: Cool.
George Wang: Alright. See you guys tomorrow then.
Sydney Bocik: You,
Lucas Menegazzo: Thanks,
David Kawata: Thank you.
Lucas Menegazzo: guys.
Sydney Bocik: guys.
David Kawata: Bye.
George Wang: Alright.
Lucas Menegazzo: Hey, dude.
David Kawata: Get a chance to go writing this morning?
Lucas Menegazzo: No. Yeah. On-It not not today.
David Kawata: I know.
Lucas Menegazzo: I skipped the the yeah.
David Kawata: I I understand. I understand. I just thought maybe that would be a a health healthy, distraction. So
Lucas Menegazzo: Yeah. And it took a few more hours of sleep today, I think, was the better option.
David Kawata: good for you. Good for you. Alright.
Lucas Menegazzo: Just give one one minute some added, please.
David Kawata: Hi, Vinny. How are you, sir? Good.
Pedro Vieira: Alright, guys.
David Kawata: Hello, Pedro. How are you, sir?
Pedro Vieira: Everything's ne, David. What about you?
Vinicius Ferreira: Good.
David Kawata: Well, thank you.
Pedro Vieira: Awesome.
George Wang: Good morning.
David Kawata: Good morning, mister Wayne.
Pedro Vieira: Enjoy.
George Wang: Hey. Okay. How's everyone today?
David Kawata: Living the dream.
George Wang: Living the dream. It is living the dream.
Charlie Gann: Minutes.
<v David Kawata>Absolutely.
George Wang: Building an company.
David Kawata: Absolutely. You know? The this is a magical moment in time that we're gonna look back to get, our choices, and and,
George Wang: Yeah.
David Kawata: the nexus is where we go, left or right. And so, it's exciting times.
George Wang: Most definitely. The future is getting created every day.
David Kawata: Future's bright.
George Wang: Future is bright, despite all the things happening in the world, people getting laid o , AI taking over.
David Kawata: Well, there there there is well,
George Wang: Future is my frost.
David Kawata: you know, it is, the bold, you know, the what does it call it? The the future rewards the bold. You know? So we are not on the beach and, although, George you know, Pedro, maybe is there a beach out there in Georgia? I don't know. What's it about?
Pedro Vieira: There is the Black Sea close to you on the left at you. Amazing.
David Kawata: There's I'm sure there's a bar to open. So
George Wang: Wherever did the laugh, that was pretty comedic timing.
David Kawata: that was perfect.
George Wang: Yeah.
David Kawata: Alright. George, it's your show, sir.
George Wang: Alright. So, yeah, once again, guys, what an intense week. We've been going swimmingly. It was making a lot of progress. And for all of you guys who haven't met Anor, there's Anor. And are you gonna quickly introduce yourself?
Anouar Mansour: That's for sure. So, like like I said, my name is Anwar. So I lead the infosec and technology at Everett. So my background is mostly in cybersecurity, but also, like, all these complex systems. And I'm just here to, like, make sure whatever is the product and experience. Decisions, they all just stick around it in, like, technical reality because I'm very deep in, like, how we can actually bring up things and make them happen. So great to meet you all.
George Wang: Alright. Cool. So, we're at day ve. Day one, we did the journey. Day two, components. Day three, prototype.
David Kawata: You showing sharing your screen? Because I don't see anything.
George Wang: Not yet.
David Kawata: Okay.
George Wang: I'm getting gonna get to it.
David Kawata: Cool. Thank you.
George Wang: Yeah. Yeah. And then, you know, yesterday, we talked about responsive design and Agent evaluation and, so great job, everyone, and truly exceptional how fast we'd be moving. And, Lucas, we talked yesterday. We know we owe your team, the mechanics of the details of how the AI, needs to respond to friction moments and, you know, like, what happens when things are loading. So we'll give that to you today, so we cannot block you from being with a test.
George Wang: Next week. And what we are let me just second. What we are, excited to share with you guys today is our attempt at simplifying the beautiful creative that our team had, had done. And I'll just On my screen now. Okay. But can everybody see?
George Wang: Alright. But before that, I just wanna highlight that. Yep. We're at day day ve of our week On-It. De ne the information architecture. Today, we're just establishing the visual directions and. It should be good for this week. And by the way, Charlie, thanks for, sharing those slides that you put together. And, yeah, we're we're bringing everything into our approach to the design.
<v George Wang>So, yep, talk you guys through the colors, typography, the photography, and lastly Some components that we put together. So, all in all, this is really just about being grounded, disarming, and candid, the three keywords that, we're focusing on. So all those mixed together, but the foundation is really grounding. So that's what you see here in this color palette. So I feel this color palette really hones in and captures the trust that we wanna have with users.
George Wang: It's modern and grounded and also at the same time keeps, me curious. And, so we have the foundational colors, and we have the secondary ones that you can choose from. And, as we diversify, how the brand will start to evolve, this will give a lot of room, you know, and space for the brand to ex and into what it should look like in the end. And we chose the font to be Switzer. You know, we thought it was important to keep it legible and, something that's trustworthy.
George Wang: And See nothing too plain that's really, ts the mark. Okay? So for photography, it's not that whether I mean, we're not committing to whether we're gonna have photos or not photos in the app, but this is really just to, you know, provide a mood of, the of our brand. So, you know, as we walk through this, we realize that for a lot of us out there, it's happiness is something that we Sort of, you know, wanna achieve.
George Wang: And now lately for us, there's, you know, happiness is reaching towards a Roam, and reach that and reach that Roam, you get happiness. But, there's layo s, there's a AIs taking over. So the last thing we would ever wanna know, is to remind people that we're AI. And so, this photography to us, are moments that are pure and are human and from hiking and seeing a gorgeous Thing. A waterfall at the end of the hike.
George Wang: You know, it's like a human really is willing to go through our little bit of hardship in order to get to that reward. And, we don't know if he did heart hike, but it looks like, you know, it it's great how he got to that but worn in the end. And here's a guy that's smiling. You like a girl looking at her window, and she's comfortable, in her own home. About her future. Is she worried about, you know, how much mortgage and how much she needs to refinance?
George Wang: So it's it's not On-It, like, gleefully happy. It's somewhere in the middle. You're kind of wondering. And so that's real. That's real to us. Right? That's that's that's why you can trust us because you understand these life moments, and you see, you know, these bags, hems and hers, and the house keys in the middle. And Yeah. It's like a grounded base where you're whether you're renting or, you're renting out multiple Airbnb units or own homes, or otherwise.
George Wang: Yeah. You see the calm tranquility of walking the stairs and seeing this silhouette of plants that may be on your balcony as you walk up your home. Yeah. You know, there's another moment of two people on the phone. It's modern. We're sharing this moment. It's like, hey. Look at this PreFi app. And, how fun it is, chatting to it. This is where, you know, we should really refinance and maybe take that money going On-It the trip. What do you think? And nally, you know, here's a lobby photo of just couple out on their vacation because they probably refinanced.
<v George Wang>So this is a, you know, the whole journey. And the point is whether you're going through the ups and downs, there there's a question mark of, whether you're truly happy or not. And we just wanna Stay grounded and be disarming. The tech is in there, but we think it's a a perfect marriage of the tech and the, you know, people in their lives, and it's not being overtaking. And lastly, as we move into these components, as you can see there's, you know, the $6.23.
George Wang: K, that's the main balance and that's the, you know, the what's left on the mortgage that, you know Like, the center of the refinance, it doesn't tell me you know, it's it doesn't like, this kind of visual doesn't not not trying to sell anything. Right? It's more like a widget that you can add to your dashboard. And so the idea is that the more information that I give to the AI, the more it these kind of widgets, the information come ll up the screen. It's like, yesterday as we're talking about the gamification, the different pieces that, information that you can add, and this is what it's gonna show.
George Wang: So it's like Getting the user to discover, to, wanna share more so they can ll up their screens with these kinds of, you know, the, widgets. So and, yeah, and, you know, as you're typing in the chatbot and you see the dashboard ll, with color and so The the Apple Watch, moving on is, you know, a little bit simple in our opinion.
George Wang: So as an Apple user, there's only three colors. There's, stand up, you get blue. There's walkie, you get green, I think. And red is where you're just walking around, moving, exercise. And, yeah, there's, you know, On-It three colors. And with this color palette, it ebbs and ows and, changes a lot. And there's a space where you can, you know, really use to capture the main demographic that we're looking for, which is 30 year old women. Leak and matches the tote bag and, you know, the Pilates studio. That's, you know, what it evokes. It's modern. It's fresh.
George Wang: And, yeah. And, with this meter, shapes, you know, we experienced a lot of shapes and arrows and, squares and just felt like too elementary and chart like. And so we opted for this, scale kind of visual. It takes a back seat and help you, help you focus, you know. There's the, like twenty eight years, rest on your loan, left on your loan.
George Wang: So, it's candid about that and, with a pretty high interest rate. These are all tough emotional things and we wanted to keep this kind of neutralness. You see how, you know, we can, dive back to be a little bit dark, but the legibility is still there, and the yellow meter is a subtle hint of, where you are. Maybe you should refinance. And now we get to this one. In in the end, we're looking at this.
George Wang: We're not trying to recreate chat, Chargebee, Cloud. We're trying to move away Roam, This is where it'll be a little bit playful, with this color palette. That's, you know, go to Roam this serious to happy and neutral and, you know, we can move either way. You just feel like you're getting, you know, guided along this journey.
George Wang: So that's, just a, you know, quick walk through of the visual direction that we're setting. Yeah. So, guys, you know, curious to hear your thoughts and feedback.
Charlie Gann: George, for the, the Swisser font, you mentioned that it's, one of the reasons you chose it was are are you using it in all your these examples here? Is that Okay.
George Wang: Yes.
Charlie Gann: And you mentioned that it it scales really well. Is that did you guys do some, just some testing as far as readability at smaller fonts? I think and even On-It just smaller, but, like Larger paragraph. You know, there's things that are headlines, which are one or two lines, which are the legibility is very different than if you're reading a paragraph. Did you guys do any sort of due diligence on that on as far as, like, how it looks and feels and reads in paragraph form?
George Wang: Yeah. Yeah. Yeah. We did. Yeah. You know, there there's things we can adjust in the type, the kerning, and whatnot to you know, if it's smaller, there's there's more space you can add. There's a lot you know, there's adjustments that we can make, and we did play around with, you know, putting a bit different sizes and see how a how they complement, how they can be complementary. So, you know, yeah, here's some examples of the Swiss Reformation action.
Charlie Gann: Cool. Yeah. So just sort of backing out, I kinda dove straight in, to a detail. Overall, I I feel really good about the direction we're taking. I really like this, over here on the left, the, the component where it's got the address and the the equity and value. I think I like that we're going a little softer, but I do think we could, like so the one on the on the left versus the component on the far right, the one on the far right is is very vibrant.
Charlie Gann: I worry a little bit about, like, legibility for colorblind and and contrast and things like that on that. It's it's it's pretty vibrant, whereas on the left Roam much softer, which I I think, you know, as we've if so we've kind of talked about our target audience, I I think a little bit of a a softer approach is is great, but I do like that where the color is on the right. So maybe there's something, you know, where it marries the two, where it's a little bit of the left, but instead of, you know, where the icons are a little more muted and darker, maybe that's where we pull in some of the vibrance.
Charlie Gann: Because I think I think color is gonna be important and could help tell Ways. But I also don't think it's it should lead the charge, if that makes sense.
George Wang: Yeah. You know, we yesterday, we talked about the legibility and, and, that it needs to be accessible for people that are older, as well. So, that's something that we we have in mind. As far as the colors go, it's, you know, we just wanna show the the different directions that it could take it. It can be playful. It can be a little bit neutral. And, yeah, there's a lot of room for this brand to evolve as we continue to, you know, especially with the What kinds of colors?
George Wang: So we just, you know, wanna create that openness and, yeah, flexibility for the, the the evolution of of the brand.
Charlie Gann: I think there's a couple I may I may have a couple more questions. Go ahead, Vinnie. Thank you.
Vinicius Ferreira: Quickly. Okay. I agree with with Charlie. Maybe, the the shot backgrounds is is a bit overwhelming right now with all the scholars. Maybe, I think, we need to, bring the accent, like, to the the widgets, those. Companies. Maybe they can maybe they are the one to to catch user attention or something like this.
Vinicius Ferreira: Because there would be a lot of conversation, I think, like, the the, the background should be, like, more subtle, more clear so people can pay attention On-It the conversation. And then when we show them those interactive components, then we can, like, use accents. Things that drives people attention so they can interact with, like because it's kinda boring just typing or, voicing.
Vinicius Ferreira: But I think we can use, like, animation, those those things to to get people's attention in the widgets,
George Wang: Mhmm.
Vinicius Ferreira: the in the
George Wang: Yeah. Yeah. Absolutely. And, you know, just for reference, here's the prototype, you know, from before. And the idea is really to, you know, take these widgets and ll it. The screen size you're typing, you know, you get this pulsating kind of, you know, here's it's it's alive. It's light. I mean, obviously, we're not showing animation here, but as you're as you're typing, you know, you see the rest of the screen populating with things that you're you're adding. You know, graphs are invisible until you add something and then it kind of pops up and you see, like, here's everything that it It's kinda like monopoly in a way.
George Wang: Right? You know, you every time you roll dice, you you go around, you got more houses, and, you get satisfaction of out of, just uncovering information, uncovering, you know, adding these switches to your to your dashboard. I don't know if that makes sense. And, yeah, it's it's about gamification and, that's the the tone that we're The feeling that we want people to have, it's, it's it's disarming. I mean, games are disarming.
George Wang: You play games because they're they're, you know, they're they're enjoyable, and we wanted to feel like you're you're playing a game as you're addressing, you know, tackling these very sensitive complex topics.
David Kawata: So, George, show show me I I don't see that. I don't feel that. So go ahead. I mean, I I get the widget. I like the widget on the left. Again, the soft pink on the pink, but I I don't like the right. I wanna be clear on my ICP on the user On a 30 year old woman. I don't want to lean into a man's thought of what a woman wants with pinks and colors. It was really the set the motif in the pictures were the greens and the beiges, something soft and elegant.
David Kawata: That that is really more what I was thinking about in the sense of aesthetically. Right? It's again, I it's I I think men's opinions of women's, what they I know you're professionals, so never mind. Would that mean? I was looking in more of the elegant version in in the middle.
George Wang: Yeah.
David Kawata: I will also say the imagery itself, again, as a clari cation, the average the median age of the firsttime home buyer is 40 years old in America according to NAR last year. The median age of a homeowner is about 56 years old. So imagery actually doesn't match the average purchaser. Right now, when I was saying that the user side of elegance and and twinnedness on a 30 year old, they just have a less tolerance for of of Are less intuitive. That that's what I really was meaning on the technical thoughtfulness of the UX side.
David Kawata: So I just wanted perhaps I was unclear. So I wanted to understand the imagery is, an actually order demographic, but we are leading with more of a mobile design 2026 design beyond. So just a clari cation if I was misleading and what my expectations were. So, just reset there.
George Wang: All good. Thank you. Yeah. We're taking notes and, we just, you know, use ping to show how you can play with the color palette and make it, you know.
David Kawata: I I love the idea of splashes of color, right, to show, moments, magical moments, excitement. I think Asana back in the day did, like, unicorns and rainbows, right, in in a very b to b white palette. So that's I I love that idea to to use the colors in a reserve but celebration. I love the teardrop kind of idea on on the components. So again, I think that there's a lot of beautiful things and intentionality that you guys brought in. So thank you for that. Back to my firstpoint on I don't see the gamification.
David Kawata: Again, I wanna provide you feedback. I was gonna do that by Monday, right, in the sense of when we could display, ideally in some type of progress or some type of missing pieces unless we get we give more value to the consumer for that clarity. Help me to understand. You were talking about your monopoly passcode. I don't see Your your, attempt or that starting point here.
George Wang: Yeah. The the monopoly, that's more of a metaphor for,
David Kawata: I understand gamification,
George Wang: you know yeah.
David Kawata: but I don't see it at all.
George Wang: You're adding you're adding, like, houses on the board just like how these components, you know, pop up. And, you know, here here's your house. Once you give us the information about your house, here's the widget that comes up that summarizes, you know, your your your mortgage. And, so as you're giving more information, more of these things might come up. And it's like yeah.
David Kawata: That's perfect. I I understand the concept we're doing. I just didn't see your example of that besides a widget.
George Wang: Mhmm.
David Kawata: Okay. Thank you.
George Wang: Yeah. This is not a full design system. That's you know, we we have the until the end of week three. To do that, this is just a very initial visual direction as to, like, the broad direction that we're heading, making it playful. We heard you on the color palette. David, we'll, you know, we'll we'll dial it back, you know, in early, you know, when we the the firstattempt is, like, we usually try to push, you know, A very creative direction that we dial it back as needed. So this is really it's all really good feedback and reactions from guys.
David Kawata: You you guys are the professionals. Again, I understand what 2026 looks like today. So, again, it was just I wanted to be clear with what I communicated, perhaps. I led you one way. I just wanna recalibrate what I said. And then, again, I'm I'm deferring to the experts. Again, mode, emotions, colors. Cool. Just wanted to be clear that that that the my my My feeling is that the users and the decision makers are generally women. Right?
David Kawata: The firsttime homebuyers age is generally 40. The women's are actually making financial decisions in greater, bunches. Single women, They're educated. They're making more decisions two to one on a single man. And then if they are in couples, they are generally the the decision makers of the in that experience by, like, 70%. Right? So the that aesthetics in the intuitive just Needs to be there, which is it is what it is. But with that also being said, it goes down to our not being pandering.
David Kawata: Right? Again, we have a bunch of guys on this call, and I know Cindy's here, but I wanted to make sure that it's not our opinion of what women want. It says, right, I think elegance and I think sophistication and beauty, I think that's universal. That was just making sure my expectations are communicated. So thank you again.
George Wang: Yep. Yeah. Please sit sit with us everyone as, you know, we're going through this process. This is the the point of this is to provide something that's, I guess, bold and get feedback so we can, you know with something for you guys to react to. This is all great. Alright.
Charlie Gann: Think something Sorry.
George Wang: Seems like,
Charlie Gann: Sorry, George. Sorry to interrupt.
George Wang: hey, Charlie.
Charlie Gann: I think something that might be a little helpful, I was taking a look at the color palette above, and I think, like, the second and third row are are really close. I think maybe what maybe take a look at that firstrow again, and maybe the combination of using some of those more muted, almost interprets a little muddier once you combine that with some of the more vibrant colors. So maybe might be worth exploring, like, either a, like, a complimentary blue or green or something, you know, that and and then some of the more base colors, something a little, little cleaner.
Charlie Gann: That may maybe marrying those two might help. I don't know. I think there's something there worth worth exploring that might help push this to towards the direction that we're we're looking for.
George Wang: Yeah. Yeah. The top is more, you know, morefineutral and the the rest is more, you know, vibrant, and we wanted to kind of show it that sort of the the balance of that in in the components. And, got the feedback today. So definitely be morefineutral and less you know, use the use the bright colors a little bit more sparingly and, and, dial back a little bit on the playfulness since our, demographic is is older, 40,
David Kawata: I'm good with play I wanna say,
George Wang: fties.
David Kawata: I'm good with playfulness and cheekiness. I wanna be clear. I I don't wanna mute that. Right? I just was thinking that that there are times to celebrate that versus lead with the party. So yeah.
George Wang: Yeah. Right. Right.
David Kawata: Yeah. So cheekiness and playfulness, I think, is an important concept to humanize the experience on something very mundane and people are intimidated by financials. So, again, I I I I wanna celebrate and encourage playfulness and cheekiness and, unexpected, moments. So I I please, I don't wanna to to sway you differently. I just my initial feedback of of the color palette was a little bit bold.
George Wang: Right. Okay. Okay. Well, thanks so much, everybody, for the feedback. So mhmm.
Sydney Bocik: Will also say too, I I like the the component of the map a lot. I feel like in so many there's so many vibe coded apps coming out now that use lots of emojis, so I'm happy to use, like, icons instead of, like, emojis here. And, like, I've been seeing them across so many apps, and it I don't know why, but it really has been driving me nuts. So I do like the map component a lot and, like, the icons, the subtleness of the icons that feel like They're not the whole point.
Sydney Bocik: So I did like the map icon.
George Wang: Yeah. I don't know if I can play this, but the map the map, we actually had a video in there. Yeah. So this is, we're we're experimenting with the API from the the aerial map tiles from Google. It had I mean, this is the White House. It's just a sample. So we're trying that with different with, residential areas. Not sure. It it seems like it could be quite slow.
George Wang: So, you know, for them to provide, you know, the videos for residential areas. So we're experimenting experimenting with how how do we make that, you know, this is a place, this is a home, idea come alive. So just, yeah, FYI.
Sydney Bocik: Okay. Cool.
George Wang: However, Making it trying to make it more, you know, interactive. Cool. Alright.
David Kawata: You, George. Appreciate that.
Charlie Gann: Good stuff.
George Wang: Alright. Yeah.
David Kawata: And this this is all in Figma?
George Wang: So,
David Kawata: Where where is where was everyone
George Wang: yeah, it's all in Figma.
David Kawata: has access? Okay. Cool. And then do you expect or would like any feedback from this today, or what would be our so Monday, real quick. Resilience, there there's a holiday for them. I don't know if that was communicated to you. We're we're I'm available. I expect us to be available, make sure that we're still on point. But if there's any expectations from the To be available on Monday, you should assume Tuesday.
George Wang: Sounds good.
Sydney Bocik: And, the other note I wanted to make that Lucas and I chatted about yesterday as well too. He said you guys had, a really great call yesterday, which is awesome. The I know the firstweek we talked about daily stand ups. If we need to reduce that for next week to give you guys more time to think, iterate, and we can work more asynchronously. I know the the call schedule was heavy this week as we were ramping up. But if you want Think or adjust on that and, shoot me an email.
Sydney Bocik: Just let me know if there's any if you wanna bring that down to, like, three days or something, if that would be more productive for you guys.
George Wang: Right. So we have the the road map here. These next three weeks, you know, we met it was designed, to be asynchronous, anyway. So this firstweek is really, you know, like, since we're all you know, we're getting going and
Sydney Bocik: Mhmm.
George Wang: Just it was important to stop, you know, like, cadence, but the next, three weeks, it's it's gonna be more asynchronous.
Sydney Bocik: Okay.
George Wang: And, yeah. We'll we'll, we'll get, you know, make sure that we we have time to to connect and and and support each other.
Sydney Bocik: Sure.
David Kawata: Real quick, with that being said, would you like to have a call on Monday, or would you like to just do it Tuesday? That's when I'm gonna be I'm gonna be more clear on that.
George Wang: Let's perhaps do times a week, and we can do next Monday. And,
<v Sydney Bocik>That would it be better that if we were to do two times a week, would it be better than to do two, Tuesday and Thursday? So that way since the team's unavailable on Monday.
George Wang: Yeah.
Sydney Bocik: Or do you wanna do you wanna check and let
George Wang: Tuesday Tuesday is good.
Sydney Bocik: me see an email?
George Wang: Tuesday's good. Let us, let let me sing back with our team on there's what the best cadence is. And, if you Know, give us feedback, today over the weekend, and, we'll make,
Sydney Bocik: Yeah.
George Wang: you know, iterate on the visual direction and everything. And, yeah.
David Kawata: Right. So, again, we'll determine what the meeting's cadence for moving forward on Tuesday. We'll just you have enough time to talk to your team. No call on Monday. And then, but we'll communicate by Roam in in email, and and we'll just keep the the momentum going.
George Wang: Sounds good.
David Kawata: Alright. Really appreciate that. Thank you.
George Wang: Alright. Thanks, everybody.
David Kawata: Alright.
George Wang: Have a good day.
David Kawata: Bye bye.
George Wang: Bye.
Christopher Everett: Hey, David.
David Kawata: How are you?
Christopher Everett: Good to see you, David.
David Kawata: Good to see you as well.
Christopher Everett: I just wanted to quickly meet, just to go over this. I know, you know, sometimes George. No. George is amazing. George is here. So let's talk let's talk about this. There's there's one thing that I wanted to do that I I didn't think he explained. And I just wanna be thorough with it. So let's, can you show the screen?
David Kawata: Think Georgia is amazing too, so I wanna be clear.
Christopher Everett: There you go. Yeah. We we love we love George. We love George. But the team was like, did you tell the client this? And I'm like, so I'm gonna help here assist on delivering this. And, again, this creative is subjective. So we are definitely not those people that we have the expertise for sure. We've done so many things that we don't wanna drop the names of. And then you I laughed when you said the the pink because it was total risk too. Like, that was just to say that we're partying. We're you know? I came from Barbie. Right? So, essentially, we can't we had to lift the demographic.
Christopher Everett: But let's go from the color palette on the top. So, like, we went through many of the the the the colors and stuff, and we really wanted something that can turn, like, really heavy, boring Re nance emotional stuffand, you know, we wanna, what, be disarming and we wanna be candid. So for me, I thought that the primary, maybe adding a little bit of green. I think that's missing. We're playing with greens and trying to, you know, avoid puke green or whatever. But I think we can blend that in, and then it'll it'll make more sense with the photography just to show you what we wanted for the photography.
Christopher Everett: But then you can see secondary is where the The components can come alive. You can utilize and leverage for some more of that pop or whatever, you know, kind of just like to say subtle pop of of semi magic. And, obviously, the gradient in the end is really where it comes alive. You see that as kinda like the soul of, of the thing. Like, you're adding life and soul to the, to the dashboard that ultimately is, you know, what I'll connect to gamification in a second. Typography is, very readable.
<v Christopher Everett>Is all sizes here, and it's just the spacing isn't too tight. The, you know, overall, we have many other fonts that we went through, but this one just really stood out for us. And then let's go to the middle because this is where I wanna really hit. So you can see the basic is like the green and the brownish tone from the primary colors, but we didn't we didn't have the green. But if you play with the green, you'll see in these individual Photos like this guy, you don't know if he's having a hard hike or a light hike or a soft hike.
Christopher Everett: Right? But we know he's staring at this waterfall, which essentially should be happy. So I think the team was thinking, like, people in this demographic, we already know that we were told, like, reach this Roam, and you will have happiness. Right? Go up the corporate ladder, and you will have happiness. So now we're our play is more of, like Moments to attract people to, you know, what we're doing, to disarm them and be candid.
Christopher Everett: Like, instead of going, hey. We're here just to help you, and we're here to refinance your house, and you're gonna get so much money if you just refinance it. Our approach is more like, we might catch you in a moment. Like, imagine trying to talk to the girl looking out the window, you know, about refinancing if you're a friend or a partner and stufflike that. So this is why we chose these photos. When you look at the Holding the phone. This is you know, he's giving love to his dog, but he's split like everyone else is. Right? We see split from, like, attention.
Christopher Everett: Like, should I give love directly and be present for the dog, or should I, look into this, what's on what's on my phone? And is that phone causing me stress? Is it happy? We don't know. And that's kinda where we want we feel like you can disarm the person, you know, that we're trying to get more information from.
David Kawata: The the imagery, again, I don't have a problem with the imagery. I actually like the color palettes. Right? So I gave you a reference in my quick note. This is actually more aesthetically. When I and I and I apologize. A lot of the I think a lot of responsibility in the firstiteration of using words like feminine or or female and not elaborating on what I meant by that. Pink was, anyways. So the imagery again on our firsttime home buyer.
David Kawata: Right? We have firsttime home buyers that are averaging 40 years old. Meaning, homeowner is 56 years old. So again, the image itself, not understand the story you were trying to create. We got busy people. They're mobile. They're Right? They're living their lives. Again, the image comment on this would be, that's actually not the person that's actually making a financial decision. Right? So I like the colors. Alice, I like the feel on the photography that you're coming out with. I would just let you know that that meeting age on a a person.
David Kawata: And then there's also life events. Right? So life events are triggering.
Christopher Everett: Mhmm.
David Kawata: So everyone on this this call is is generally the people that grew up that the financial crisis actually had effect on their families, generally negative.
Christopher Everett: Yeah.
David Kawata: Ownership and the, and that is actually being questioned. Now they're getting into a stage like they're almost running out of They're adulting to a point that's so hard that homeownership now makes more sense. And so that's the only comment I would have on the photography that I didn't put in my note is that the imagery is that they're actually older in the decision making process, and their life events that are actually triggering everything that they're chasing, running away from home ownership is just becoming a reality.
Christopher Everett: Mhmm.
David Kawata: So, again, the imagery was be more on a wellness kind of lifestyle. App. I say that that is wrong. I do understand the messaging of the imaging and I like the color palettes in the photography. So that would be my feedback. So again, I, I, I never questioned your storytelling or your softness or your awareness
Christopher Everett: Great.
David Kawata: on the empathy of the imagery. That was just my feedback.
Christopher Everett: Sweet. Perfect. And then, and I'm happy that you get you keep giving really good feedback. It's it's super helpful and never feel you know? But I always tell the team, like, never marry to anything and be like, oh my god. You won't take it. So this is really helpful. Thank you for saying that. Can you scroll down, George? So this is the, where's that prototype, the black and white prototype?
George Wang: Right here.
Christopher Everett: Yeah. So, I mean, this is just un nished. But, if you can ignore, like, the colors and stuff, what how I'm seeing this is when I think of gamification, I know you're gonna give that to us on on, Monday for more clari cation. But for us, when we were thinking about this is it all starts with the chat and the conversation. And so as we start to talk to them, imagine, like, the widgets and the components going into the dashboard, and that is the Ultimate reward, you know, the the personalization milestone.
Christopher Everett: It's how we see it. So you wanna give to them more information, like, oh, let me put in my address. And then Roam. All of a sudden, you're getting, you know, a shot of your house, a video of your house, whatever, and then you're getting the, thanks for the compliment on the icons. Sydney, I agree. Too much emojis too. So we we we the Roam stay stays far away from that. Everyone's just copying everyone. But then you start to make this dash dashboard come alive. The more you, you know, talk to it and and, that's literally what we were trying to to hit.
<v Christopher Everett>Did I get the points right,
David Kawata: Yeah. So,
Christopher Everett: George? Okay.
David Kawata: so, again, I I think that goal is absolutely correct. Right? The more you give, the more you get, the more you get. You want to give more. Right? Create that positive experience of, giving the giving. Slight heartburn is that from a mobile firstperspective, that is the constraint challenge that I would again, I wanted to, emphasize, and George has acknowledged it. Again, acknowledging, where we need to be mobile first.
David Kawata: And so I would I would challenge that when we create visual stories on accomplishing the awareness and these, this, enlightenment moment that we, we see how it actually lives in, in mobile, because I think That would be the harder components versus all these beautiful components we're building, and then we're trying to figure out how it works. And I know you know that. I'm just reiterating from my perspective from week one. K. Two,
Christopher Everett: Yep.
David Kawata: I thought that that again, I'll be just very, candid, because that's one of the words. The the pink was disappointing as as a one that you came, not because it was showing energy or fun. Right? Because I took it that when I said that the ICP was a 30 year old woman. With pink. And I actually was, that was disappointing when I firstsaw it. Just gonna communicate that to you.
Christopher Everett: Yep.
David Kawata: That that was, a lazy, response. Not saying that was intentional, but I'm just that was my feedback. Second, on the research, I know that on day two, we were talking about market research and other competitors. The tracker from the the middle component, no one shops in in small increments when you're talking about hundreds of thousands of dollars for Roam mortgage. So it just The translation from research to components to the visuals, there was just slight I wanna call it sloppiness, but didn't didn't translate.
David Kawata: Right? And then, of course, the third component when we talk about gamification on the desktop did not seem to be aligned with our needs to be mobile focus. So those were kind of my three harbors on the three components, initially. Again, George has been very clear and communicative that we know that this is mobile. Solve the mobile. I know that there's a plan, Chris. Christopher.
Christopher Everett: Right.
<v David Kawata>Excuse me. I got it. I also know that, the the reveal and the gamification is not been shown, but it was talked about. And so that was another when I saw the components, that's a that's a desktop component and I and that's a that's a, like, a piece, right, versus the reveal. So I did not feel that the story points were illustrated here. And so I've just that was my initial feedback. Roam the my design brief that I gave to you that I might have been misleading and what we're getting back.
David Kawata: Right? So I I give you and I'll put
Christopher Everett: Yep.
David Kawata: it better in email if you prefer. But a quick little text was, again, not that it's a thirty year old ICP, is a thirty year old that the UI UX needs to be beautiful and too intuitive. But they're not actually thirty years in the decision process. They're actually closer to forty years. On the firsttime home buyers and they're actually older.
Christopher Everett: Yep.
David Kawata: And it's actually life events that are triggering me. And I don't wanna be patronizing and to to someone at any age that we're trying to make this feel like a wellness app. This is a large financial transaction, and and we wanna be respectful in
Christopher Everett: Yep.
David Kawata: that sense of expertise. And the last point, again, George made a point using the colors for fun. I wanted it to be fun. I wanna be personable. Again, I wanna be less serious in the sense of more of a Right? Not not, not loose or casual, but more woody and cheeky. And so I put that slightly in in there. And what I meant by magical moments, the magical moments are not just the witty comments, but using these burst of colors or animations to to show that And that's again, I don't know if I articulated that well the firsttime as well.
David Kawata: So that's on me, and I thought this was just a really good chance to be. Expectations because it is subjective, Christopher. And I and I know that, and I did a bad job, I think, leading that up. So that's kinda where that's my firsttake on the deliverable.
Christopher Everett: Great. No. I love it. We really do love it. We we we sometimes people, Agency I know, like, block the initial feedback because it's like, let it sit. Let it but as you can see, we like initial reaction first, and then after that, looking forward to the email. Yeah. Email, and we're taking it right now too. So all this just helps, you know, this project go along.
David Kawata: Perfect. But, again, the thoughtfulness, again, I I I, again, I have no doubt of the thoughtfulness of the team. I again, the middle component does show that the research or the knowledge is didn't get translated into the into that, which is, my feedback to you that if there is gonna be delivered, if we could be aligned with what you actually mean, I would appreciate that.
Christopher Everett: Okay.
George Wang: Thank you, David.
Christopher Everett: Thank you.
David Kawata: And and else do you wanna, again, get clarify on the original brief, my notes back to you? Is there, do you think that any of my feedback is unfair? I don't know if unfair is the right word, but that you wanted to perhaps unpack some more so that we we we are more on the same page.
Christopher Everett: No. I think, this this slight not bad. The slight, course correction is is is a good one. It's a good time to to do that. So I'm I'm looking forward to taking, all this and, you know, I think we have enough. Go on.
David Kawata: Yeah. And I know they were course o . Right? I think it's just a calibration.
Christopher Everett: Calibration.
David Kawata: And yeah.
Christopher Everett: There you go. I I just can't see any adjectives today.
David Kawata: I think so I I get yeah.
Christopher Everett: I blame it on the kids. That's it's it's all it's all about kids. You know?
David Kawata: Oh, the sugar of the kids. Absolutely. So, again, again, I think I I misled in some of my verbiage in the brief. And so, again, I appreciate you giving me the space to recalibrate and clarify. I'll I'll follow-up in an email for your documentation so that you have that as a source outside of Roam.
Christopher Everett: Great.
David Kawata: Sydney, any additional thoughts, feedbacks from your side?
<v Sydney Bocik>I don't know that I have anything to add. I think the the, the website you gave today, I think, was really interesting. I had I hadn't seen it before, the Andaman Resorts as as well.
David Kawata: Mhmm.
Sydney Bocik: And I like the have you guys seen the that one from, a design perspective, I like because the clean and white, I think it articulates what he was communicating well. And did you guy I know it was kind of buried in the knowledge transfer, but I like the the component that we're bringing up here that,
Christopher Everett: Yep.
Sydney Bocik: because it kind of reminds me of origin, the use origin that was in the knowledge transfer that, has always been in terms of, like, components and messaging and the warmth or the tone of the website has always been something that I really liked and had come back to. So I think an So I see some of that coming through in the in the components here that you do have. I know we're talking about some adjustments, but it's, exciting to see that come in.
Christopher Everett: The team will definitely smile on that one because, origin was one of our, you know, reference On-It on what we liked.
Sydney Bocik: I I like their website a lot and the components.
Christopher Everett: Yeah.
Sydney Bocik: And it didn't feel,
David Kawata: Perfect.
Sydney Bocik: something
Christopher Everett: We have some. We just don't wanna, you know, we don't wanna store cards yet. Just kidding. Our reference list is pretty, is amazing. I can't wait till it comes all together.
David Kawata: Perfect. Again, and the other the other thing that I would just refer, and I know George has already acknowledged my comments this week. Again, readability, ease of use,
<v Christopher Everett>Mhmm.
David Kawata: font sizes. Right? Again, I don't know your team, but I'm gonna assume there's there's a lot of males in their twenties.
Christopher Everett: Yep.
David Kawata: That's just how I'm taking this this initial feedback is that they're males thinking what women want and people with young guys and people that are active and and that's how I've seen the designs. And that's actually not gonna be probably our real users. And so, again, if I'm wrong, I apologize, but that's my initial take, to really understand the usability of our users. So We could be sensitive to that or at least in a way for the users to change the I know George talked about the device as ADA type of accessibility type of features for someone that has extreme problems or someone of a certain age that already just says everything's wrong for them.
David Kawata: But I wanna be mindful that the designs are tight and small, and and they're young in the in the sense of design element. And Classes that I can't read menus. So I'm just trying to be sensitive for for that as well.
George Wang: Yeah. Absolutely. Yeah. You'll see that in the next iteration, that we
Christopher Everett: I'm there with you, though, on the glasses, by the way. Yeah.
David Kawata: You're just trying to make me feel good.
Christopher Everett: I was just no.
David Kawata: That's but I appreciate that.
Christopher Everett: For Roam like, with the optometrist told me, hey. Bifocals are good now. You're you're kidding me. Right? And they're like, nope. You're you're you're getting close. I was like, great.
David Kawata: It's unfortunate I don't need them, like, most of the time, but just when I read and the problem. But, anyways, I appreciate you giving me the space for the feedback. Again, I know that we're in good hands. I know that you guys are professionals, and this is a subjective process. So, we're we're all just trying to mold that clay on the on the on the stool. So, again, this is also just my my initial feedback.
Christopher Everett: Right.
David Kawata: I will have few other people to try
Christopher Everett: Mhmm.
David Kawata: to Give me their perspective as well. But this was I think I just misled on the brief, and I wanted to communicate that. So thank you for coming on the call so quickly.
Christopher Everett: No problem.
David Kawata: Anything else, George?
George Wang: That's all. Appreciate your time, David.
David Kawata: Good. Anything else?
George Wang: And, yeah. And we're, we're working on some of the deliverables for, for the big Roam to get them blocked on how to address some of the friction points in the conversation. So, Yeah. Like, we'll probably we'll try to send it today, and, we'd love to get your feedback on that as well.
David Kawata: And then Christopher's already acknowledged that I have a deliverable for you on Monday. Again, for the note taker, the goal there is when there are certain data that's been given to the from the consumer, what could that unlock? For example, the property address can create that nice component on the left side as an example. When we get into, giving them a scenario benefit, right, can provide a new card.
David Kawata: Right? So these are some components I'm trying to think about the framework for giving them clarity on the ability to say yes. So that would be my deliverable for Monday. Okay?
George Wang: Great. Sounds good.
David Kawata: Okay. Yes.
George Wang: Alright.
Christopher Everett: Alright,
George Wang: Thanks,
Christopher Everett: guys.
George Wang: Paul.
David Kawata: Thank you.
Christopher Everett: Have a great weekend.
David Kawata: Appreciate you. Bye bye.
Christopher Everett: Bye.
Lucas Menegazzo: I said, because we just pointed out my my bed wasn't the the best as it could present herself, so I'm just hiding it a bit.
Sydney Bocik: Hi. I didn't even see it. Sorry. It's okay. I think we're all working in our room. My bed's, like, right there.
Lucas Menegazzo: Yeah.
Sydney Bocik: That was it's starting to nd,
Lucas Menegazzo: I think I can, like Again,
Sydney Bocik: like, my bed my bed to my desk.
Lucas Menegazzo: in the way.
George Wang: Hey, guys. Good morning.
Vinicius Ferreira: Morning.
Lucas Menegazzo: Hey, George.
George Wang: Actually, it's not the morning. It's morning for David.
Lucas Menegazzo: I have the MCP running up, Sid. It's now connected to the conference.
Sydney Bocik: Very cool. I know I got it connected to Notion, so it's been awesome.
George Wang: Oh, you gotta connect to the Notion? Nice. I'm gonna try that as well.
Sydney Bocik: Yeah. So all of our meetings at the end of every day it's really nice. It takes all of the meetings from the day. Puts it in Notion. Creates the all the decision logs from every meeting we've had in there are now in Notion. All the to dos are also sorted. So, like, that happens on a daily basis kind of automatically. So so,
George Wang: Great. Great.
Sydney Bocik: Lucas, we can also chat about moving stuffinto con uence and splitting where that split happens, but just so,
Lucas Menegazzo: Mhmm.
Sydney Bocik: all the decisions and everything from at least the stand ups are.
Lucas Menegazzo: I see.
George Wang: Cool. Alright. Welcome to week two.
Sydney Bocik: So No.
George Wang: First of all, thank you so much, David, for David is in need for your feedback. On Friday, with the, on Monday, sorry, with the, brief for, the give give and, get brief. Super, super helpful and really Know, focus our team's, you know, efforts on, achieving that north star. So what I wanted to do today is to show you guys an updated wireframe and prototype of what we got to and get your thoughts on a decision that we've been wrestling with, around the scenarios.
<v George Wang>So I'll just share my screen. Okay. Right here. Alright. Keep in mind that this is a wireframe, and, we're starting mobile mobile first. And, this Friday, we'll show a more updated, visual register, and next week is when we really Focusing on the, design system.
George Wang: So this is work in progress. So as you get to the, the the the product asking you for your home address or you just have a way to ask about anything without necessarily putting in your address. So let's say I enter my address. It's gonna pause. Right. No loading. It's just, taking a beat. And, so he's like, one of the challenges that, you had outlined, David and Sydney, is how do we how do we kind of surface, and integrate the Clarity report so that it's not sort of it's it feels seamless.
George Wang: It's not fully you know, it's not this Other thing that has to be, that's separate from the conversation and and how do we integrate in a way that's that feels natural. And what we ended up doing is using these little, data pills that, basically, has used surface information or has the AI posing information from the APIs, whether it's Roam or other sources.
George Wang: It surfaces these insights. Here it says equity, and there are different colors. So yellow is, you know, your wealth. Green is safety, and pink is, like, it's, you know, insights. So there's different color coding for, you know, different kinds of information that it's surfacing that will eventually go into your query report. And so it's not something it's not super in your face. Right? Like, as you're Giving information, it's like it's it's it's it's On-It showing a progress bar.
George Wang: It's like there there's these little dots that ow down from when when the information is provided down to the bottom. It's very, it's there and you so, you know, providing this, Feeling of I'm accumulating knowledge instead of it being extracted from me. And as you can see, these are the examples of the the drops that you had outlined, graphs, graphs, so as I'm continuing, here at the bottom, it says your picture is taking shape or it's saving before you close this tab.
George Wang: Inviting people to save, save their progress and the little dots comes back to declare your part from the bottom. So it's all, it's all connected. And, save my picture is where I enter my email password, and user knows what gets saved. So it's, they're always reminded of this is being done for you and, save back to the conversation. I can continue to ask, about, you know, different scenarios and, you know, these kinds of drops might come up from time to time.
George Wang: Here's a example of a calculator type of, drop. And this is how we're, you know, thinking about scenarios so far. So There are three options available to me. If I click on one of these, I can see a more detailed view. The the question here is that inside, the PRD, it said, the scenarios should be tailored for each of the archetypes.
George Wang: So optimizers, they, you know, they they wanna see the numbers. They wanna they they have a very different way of looking at this process than somebody, you know, looking for. Relief in their monthly payments. And the the the content of these scenarios should reflect what these what each of these, user types hold dear. Right. So first, I just wanted to, surface that.
George Wang: And, you know, skeptical, you know, we're not choosing anything for you. It's just path a, path b, path c. Just you're not being led anywhere. First of all, this way of tailoring the content, the title of each scenario based on archetype, That's still what, what we think is the the best approach.
David Kawata: Alright. I think I think, everyone here probably has a different way that they like or they like to learn. Right? So that's that is where we would hopefully be able to, respond to archetype. In in that way that again, that would be the magic to make sure we pick it right to communicate it. But I I think that, yeah, I think some people like text, some people like conversations, like to be told, and some people like to have the the information. So That concept,
George Wang: Yeah.
David Kawata: you're you're you got it exactly right. So, again,
George Wang: The the degree to which these things vary, that that's something maybe we can, we can play with. So is it the titles that's like, even the you know, all the scenarios are named differently. So as somebody move from an optimizer to a skeptic? A skeptic mid conversation or cash ow, you know, should the names change as well? Because if that does, you know, the name change, then it become, you know then do do I recognize these three scenarios anymore? Right? So if the names don't change, then are we just changing the things inside?
George Wang: So it's just the negotiation that we're, you know, we're we're doing. And, what do we x? What do we change so that it's tailored without it being They changed sort of archetype. That make sense?
David Kawata: Does. And and, again, I think math is generally, people run away from the concept. Like, they're like, I'm done with school. I never wanna see numbers again. So I'm just,
George Wang: Yeah.
David Kawata: it is a challenge on being helpful and versus making it feel like there's homework. And so, that's, I think, some of the the magic that we have or the the tightrope that we have to walk. Right?
George Wang: Right. So what we currently have is this type of, sort of overview on the key pieces of data. So if they're these, you know, key numbers, they'll change depending on the archetypes of, you know, payment ceiling for the risk reducer, monthly delta for the skeptical breakeven. It's all it's all somewhat different. Below is is is, you know, similar.
<v David Kawata>Is there is there, like, just a thought about, like, an accordion type of an extension where there would be some information, but then more information is opted into process. Right? So what I what again, it's clean. I think you've done a lovely job within just a short period of time. So and you've been you've been responding to mobile. So I again, I think you've listened to a lot of our feedback.
David Kawata: Have you guys played around with the concept of an extendability for the visualizations, the graphs, so that it's more on demand than just here? And have you played around with that concept?
George Wang: Yeah. We're thinking about, you know, maybe putting each of these parts in its own card, and you can kind of cycle through them and see different parts of the, you know, the the scenario. It's something that we're, you know, we're playing with right now.
David Kawata: Because I just think that a lot of times over again,
George Wang: Yeah.
David Kawata: for for me, I'm I'm okay with graphs and numbers.
George Wang: Yeah.
David Kawata: Like, I process that well, probably better than text, but, I think that my initial thought is that it's a lot of information, and then I have to understand the ow. Like, what does this mean to me? I live in numbers in this one topic. So but I don't think that this is More of, an option for the reveal from a customer user side of it, where here here is here is the tidbit, the reason why you may want to go, and if they could have a kind of open the door if they want to or they could skip that pathway.
David Kawata: Right? So I I think, I'd love to have you guys think about that in the sense of how can but, George, you talked about your your mother, your grandmother having, like, have one technical Option to technology, right, versus perhaps your your, someone that's not a consultant, right, more of the art side. I love to see how you think about that, that doorway potentially versus here here's the the road to Nirvana.
David Kawata: Go. That's a long road and might be scary. So something in that love to see how you guys think about that.
George Wang: Yeah. Yeah. So ease into it a little bit more is what I'm hearing.
David Kawata: Where the customer could actually kinda tell you which they want. They want the graphs. They want the numbers. Or they just want like, is there a way that they could pick your own venture where they're actually raising their hand more explicitly versus us to try to get the tone for the archetype.
George Wang: Right.
David Kawata: I think that will allow us the user to feel more in control, and then skip what they don't want. And I think that's as important. Right? It's, we wanna be helpful and be really smart, but as Jarvis only kinda helps to answer the It's a quiet intelligence. This is just showing how smart we are, and I think that there's, I'd love to get your thoughts on that.
George Wang: Yeah. Absolutely. I think I think right now, it's we're just showing you an overview, and then you can and then there's the match,
David Kawata: Oh, no. I get it.
George Wang: and then there's the next moment.
David Kawata: I wanna have all the components so that the actions there
George Wang: Yes.
David Kawata: I get perfect.
George Wang: Yeah.
David Kawata: But, but I wanted to just, give you.
George Wang: Start easy and then ramp up and give the user choice to it's like a progressively ramp up if they choose to.
David Kawata: Voice. Yeah.
George Wang: Right now, it's like, if you struggle, then we'll give you something a little bit simpler, but you're one you're looking for, you know, the other way around. Yeah. K. And yeah. And, you know, the other the the thing about, these sort of contradictory, requirements of archetypes and and, the naming and the, you know, the content of these cards and how they're tailored to different archetypes.
George Wang: That piece, how like, we're trying to think through, and so I'd love to get your thoughts here. How does that work together with this person could change an archetype halfway through the conversation, maybe after they see the scenario and they're, you know, they're a different archetype now. So, like, did the name change? So there's a little bit of tension there. We're so either, you know, we don't detect the archetypes, like You know, we've decided there are there's archetype and then we don't change it, or we, don't change the names as much because otherwise it would, you know, people would these, you know, they would kind of lose track of the the scenarios that were generated for them.
George Wang: Does that make sense? Like, what's what's your take on that?
Vinicius Ferreira: Okay. George, something that I think it can help to understand, like, people's motivation, it's in the very beginning, like, showing some cards. I don't know. Something where people can easily, understand, or just, like, drives the system for them for their intention. For example, I want to Re nance. I know. I don't know. But, this is something, I believe it it can helps to to drives, the system to, their archetype.
Vinicius Ferreira: The system needs I don't I don't know. I don't know. Like, mid the the path or something like this in the very beginning, and then going to the the address, maybe that that's a way to Yeah.
George Wang: Right. So have the user choose, you know, what their But they're yeah.
Vinicius Ferreira: Because just just, giving the user just a blank space is too much. But if you give the blank space, but also some kind of options, they can have a, like, both. Like, they can easily scan, the things that the the system can does, can do. And, and also, like, just go through the conversation, something like this. Also something that I I think is very, very nice here is when you have, like, the scenarios, you were asking about, like, like, different archetypes.
Vinicius Ferreira: They they have, like, different preferences. Maybe after, the system de nes the archetype, they can okay. We can have, like, different ways to demonstrate the data. But, also, maybe we can have also something, here that can make the user select, like, what kind of I want more data. I want less data. I don't know. I'm I'm just, like, brainstorming here.
Vinicius Ferreira: But mhmm.
George Wang: Yeah. Yeah. Yeah. We can we can definitely Roam more on that, but it's like it comes back to this. I think there's in the PRDs and there's there's this there's, sort of focus on detecting the user state and constantly adjust how we approach them based on the real time state that they they they're in. And meanwhile, you know, we wanna sort of keep, you know, not not adapt too much so that, it's like the names of things that they're looking at is all of a sudden different.
George Wang: So and and that applies here and apply it's like a more meta type of, you know, tension that we're observing.
David Kawata: That's right. I think this more nuance. Right? But I think the customer, if we allow the customer to help us, help them,
George Wang: Ja.
David Kawata: right, would would be probably more The outcomes would be more predictable, right, initially as we because we're all learning our training and other stuff. Right? So, again, if we invite the customers to But, yeah, I think it's the right direction. I think I think you you're listening in the mobile. I think there's a the, like, a light like, a huge jump just in twenty four hours from the Monday feedback.
David Kawata: So I wanna just give you that that notice, but I just wanna welcome the the idea of not just making sure that we're the smartest person, making sure that we actually know the architects to be the a 100% right on the prediction. But these cards and these There'd be graphs, text, videos, what we can generate now with AI that we couldn't do six months. Like, can we can we support the consumer as they're opting into the ow? So we get their fastest yes or fastest clarity moments. So I think that's just the the
George Wang: Yeah. Yeah.
David Kawata: the the what I'm trying to make sure that,
George Wang: Okay.
David Kawata: we're we all wanna be really smart people. Engineers wanna make sure we we optimize, you know, the process, but sometimes the process just needs to be Like, we don't need to optimize something that doesn't need to be there. So if a customer can go ahead, Sid.
Sydney Bocik: Two. And I like, I like the namings that you Even if they do change or it takes a little while to figure out what kind
George Wang: Okay.
Sydney Bocik: of their archetype is, the the naming on these I actually like the naming kind of style and concept on the other ones because, like, path a takes up a lot of space, but unless I know what these numbers mean, it doesn't really tell me anything. So I like the idea of, like, coming like, naming them something like this even if but then sticking with it even if the archetype changes because it might take a little while to Figure that out as well too.
George Wang: Yeah.
Sydney Bocik: But I think this is, like, this this makes this makes sense to me. I I think I like this a lot. What it what is it doing for me? And then breaking it down by, like, how does that how does that amount? Like, this right here, I think, is great.
George Wang: Okay. Right.
Sydney Bocik: But but I do agree with the concept of, like, if their archetype changes, we have to keep we have to keep the naming conventions through the through the conversation. And then I don't know how that works from, like, a like, a back end data perspective in terms of maintaining that. But,
George Wang: Yeah. I mean, like, we're, you know, we're imagining people get to this part pretty quickly. Unless there's a lot of, you know, like, unless you're having a very long conversation with them, being able to correctly figure out which archetype they belong to, in just two or three sentences? Maybe. And but, you know, that there could be errors there too. Right? Just because you don't know them.
Sydney Bocik: Yep.
George Wang: And so I'm what I'm hearing is lean more towards having them decide, and the archetypes can play a role as well. But giving them you know, letting them pick the pick the path that makes the most sense to them.
Sydney Bocik: Because I think that if the architect comes through more on, like, the way we're explaining, then there might be more wiggle room in how they respond to gure it out later versus, versus, like, making it come through in all of the naming conventions for everything. But, like, the the explanation sometimes the explanation might only be, like, a sentence or two or a tag or something. But Through on there.
Sydney Bocik: Maybe that helps it simplify it, but then keep the consistency. Because then the naming conventions won't change, but maybe the that's when, like, the the explanations can adapt as the archetype changes or if
George Wang: Mhmm. Yeah.
Sydney Bocik: we're just not sure yet.
George Wang: Right. Right. Makes sense. Yeah. Vinnie and Charlie, maybe we can get together and, you know, have them Roam on this and try to figure out what's the what's the best way to to guide people onto this journey.
<v Sydney Bocik>And I but I do think the the mobile view is great. Obviously, there's always I think I think the the direction is, really exciting to see. I think also too there's always ways and David can and I can help to probably figure out what information to skim back, that wouldn't feel overwhelming. So this is this is exciting to see. Thanks,
George Wang: It's it's a small screen and and
Sydney Bocik: guys.
George Wang: especially with older audience, accessibility. It's it's, you know, it's, prioritize prioritizing is is important. Yeah. So,
Vinicius Ferreira: Also like the way you you pin the data. You you can easily review it. This is I found it really clever.
Sydney Bocik: Yeah.
George Wang: Yeah. So I mean, this is just a different different way of entering.
David Kawata: Also did like the entry point, that firstdoor, because you gave them a bifurcation to have a conversation to skip the property address, but the property address was the lead. Right? It was the unlock. So I thought that was a really nice invitation that looked natural to prompt what we really want to really help them, but allow them to with a without a lot of friction to go into a conversation. So I thought that bifurcation front door, was a nice
George Wang: Oh, thanks. Okay.
Sydney Bocik: No. Vicky, I agree with Vicky's point too here. This the thing that I nd sometimes, like, chats get so big too knowing that, like, it's kind of keeping track, but that it's also building intelligence on me to use maybe for bigger insights. Like, I I agree with Vinny. I like what you guys did here a lot. And then now I now I know that, like, I don't have I'm not gonna have to scroll forever to g figure out, like, what you said.
George Wang: Right.
Sydney Bocik: That annoys that annoys a lot of me sometimes. Look in chat. So Yes.
Charlie Gann: Building on sorry, Pedro. I just wanna add to Sydney's real quick, because I agree. That was one of my favorite pieces. However, it took me a while to notice it, and I didn't understand what it was at first until you interacted with it. And I'm just thinking about this on a mobile device. If it's within a browser, it's kind of getting crammed into several things. So may wanna consider, like, how do we make that a little more prominent, obviously, without taking over the important content? But is there a way that we can Firm, like, this stuff up here, we're saving it in in case you need to reference it.
Charlie Gann: But, yeah, that it's it's a really cool feature.
George Wang: Right. Right. Thank you. So Mhmm.
Pedro Vieira: And just just expanding on that on the appeals guys, you we're actually solving a problem that might be didn't didn't expect, which is making sure that we capture the right information. Because the user is gonna say, oh, I have, like, $20 or they're gonna use some kind of, terminology that is non obvious. Right? And the AI is gonna pass and extract that information. So by showing them to their here's what I captured. It's some way more elegant way than, like, prompting a a chat with, hey.
Pedro Vieira: Alright. So I'm not sure if you thought about a mechanism also, George, to edit this kind of information because I I don't know, like, platforms that I go. I'm really big on simulating different scenarios. Right? And I go on and change and see how the the the thing adapted to it. So I love the concept. I think it solves also technical problems on our end, and just provoking you in a bit in in terms of well, The user be able to edit this information?
Pedro Vieira: How would that blend in with users that might be exploring different scenarios, for example?
George Wang: Yeah.
Pedro Vieira: Another thing that I like to do is especially in financial things, I like to put in my scenario, and I always think about my parents. They don't have much of a financial education, and I usually try to simulate their, how how they they are their scenario. Right? And see how the the two behave. So how this kind of, feature will behave when we have Multitude of information, that's something else to to take care. But great direction here.
George Wang: Yeah.
Pedro Vieira: It's really good.
George Wang: Yeah. That's good, man. We're designing this for our parents, David included. So, yeah. Great. Thank you, guys, for all the feedback.
David Kawata: General favor. Of of the language of the month. Oh, boy. So, month fourteenth, I I for Roam grammar's sake, I like in fourteen months. Right. So just that that would be my word preference. You said month fourteen three times in this this snapshot. So by month 14, breakeven month 14, month 14 versus, in fourteen months. On-It know if that's really, as just a personal issue, but, continue, please.
George Wang: Yeah. But the right. But the the the way that these drops look and feel, they're they're in, you know, they're they're okay with you. That's what you're thinking about in your head.
David Kawata: Yeah. No. So, like, you did three different things with that. Right? The blue, it's text. You have the green kind of the summy some summation of it and then the visualization of how they relate to the cards and to that point on the archetype. One of those Three may be the way that I want to be communicated best, easiest, and what I was expecting, but I have the ability to deliver in a way where Charlie gets it really quickly or Vinny really gets it. Now with that being said, I don't wanna show all three as a natural default, right?
David Kawata: It would be as and right that to me that those three things are actually all redundant on the same outcome but they're all necessary for the archetypes that we're trying to make sure that Christopher gets what he wants and he feels like, I gotcha. Right? So that, that was all I was trying to say on the archetype. It's, it's meant to make sure that each one of us is getting what we want. So our moments get there faster, but not necessarily presented in a way where it's just all the doors.
David Kawata: Because I think that now can becomes overwhelming knowledge. So that, that would be the, the balance. And that's where the pick the ventures see how they're continually creating the habits of picking this type of This way that I like this or playing with the calculator or skipping the calculator. This is where The US side of it, I think comes in back into is that, okay, they're skipping the graphs and they're they're responding to the text and they're just chatting. Then that archetype defaults into the way they want to be responded to. That's how I was kinda imagining the archetype visualization cards kinda playing back and forth.
David Kawata: But, these are just assumptions until the customers tell us.
George Wang: Okay. Okay. That makes a lot of sense. Okay. Alright. Thank you so much, guys. We will incorporate the feedback into, into the next set of iterations,
Sydney Bocik: No.
George Wang: and we look forward to meeting again on Friday. Alright. Have a good day.
Sydney Bocik: Right. Thank you.
Lucas Menegazzo: Bye bye.
George Wang: Bye.
David Kawata: It's like I just saw you.
Pedro Vieira: Yeah.
George Wang: Hello?
David Kawata: Hello? How are you,
Pedro Vieira: George.
David Kawata: sir?
George Wang: Man, lots of work. Lots of work. How are you guys?
David Kawata: Well, thank you. Well, I think, I think my comments were at my 12:30 in the morning, and then I know Sid works about 02:30 in the morning. So, yeah, we, no no time for the weary.
George Wang: Not On-It a lot not not a lot of sleep, combined here between all of us.
David Kawata: Yeah. We could change the world. Does it you know? Change the world. We gotta create our day in the universe. So
George Wang: Exactly. Exactly. Oh, I'm really excited to show you guys, where we got into so far. I mean, things are still in ight. So you're gonna see, you know, a lot of, a lot of new things, but, you know, bear in mind, everything's still a work in progress. Sydney gonna be joining us soon.
David Kawata: She'll be she'll be here soon. Did you get to I think you saw some of the comments. I I saw you, navigating some of that. Hopefully, you had some time in advance. So great.
George Wang: Yeah. Yeah. We did. So, I mean, there were a couple of different work streams that we have design team that's, you know, working on just pure Figma. And then, you know, some other people were trying to take the Figma and and the logic of the agent and marry it together. And then there's all these pieces. How do they how do they connect? Like, scenarios, clarity cards, the conversation, and try to prototype it out so that, so that, you know, we're we're not just, like, designing these things in hypotheticals. See how it all works.
George Wang: And so so let me just, share share my screen. And yeah, let's see. K. So now you see all your faces. Sharing two screens here because On one hand, it's the Figma. And on the other hand is the, what's the what's the what's the shortcut for for, getting rid of the there you go.
George Wang: Alright. So let's just go through the Figma first. Actually, you know what? Let's drag this over here. Let's go through the Figma first. Just wanted you guys to get a sense of the visual direction that we're heading in. Right? Like, before it was that, like, This, sort of color we added a little bit more, you know, refinement, a modern touch, make it, you know, change the the fonts. Right now it's, rethink. I think that's what the team had, you know, rethink.
George Wang: And then, inclusive. Inclusive stands for the body. Inclusive is I mean, as the name suggests, it's the most inclusive On-It. And so that being Inside, you know, that it's almost like a, like a way to position yourself. It's like we're so we care so much about, you know, ensuring that everybody is included, that, you know, we chose the fund that that represents that ethos.
George Wang: And, actually, you know what? Let me let me just share the Figma first. I think that might be easier than that. I can turn over turn over into the prototype. Okay. Alright. Here we go. Okay.
George Wang: So, just focus on the visual elements. Right? And then I'll show you the portal type. We'll have some q and a around, you know, the different options, what do we prefer. So yeah. I mean, this is the color scheme. This is the the one that we're we like the most internally. It's cool blue. It's calm. It's like, the lake, the water. And, you know, the the Roam advisor having this sort of futuristic kind of, avatar that While it's, talking to you, you have these pieces like what is refinancing when there is, some terms, you know, inside the messages that people might might wanna learn more about.
George Wang: And so it's giving, you know, play paving a path for people to who are not as familiar with with this world to to on Roam themselves, addressing your feedback, last session, David. Drops. This is how, you know, they'll look like broadly speaking gray, white, the, the login, we're thinking of putting it in a different, making it slightly muted so that as you're navigating through this, even if you're, you know, if you're not we don't want we don't want that to be shown all the time.
George Wang: Right? Like, you know, just again, just trying to make it not feel like extractive and and salesy and you have to log in, like Be a constant reminder of that. And so we'll make that a little bit muted, but, yeah, it's the overall, you know, color scheme and feeling that we're trying to convey in this in this design. I know, Vinny, that you had made a comment about, you like, putting these these, little pills, underneath the messages.
<v George Wang>And we tried that. It just it just lost the It made it like, it took away it took away some the color that makes this, this interaction feel, exciting and, and interesting. And, yeah. You know, this is what we're thinking about for the magical moments, changing the colors so people can, users know that something's different. And when you actually see it on screen, we don't have it wired up wired up in the prototype, but right now at least.
George Wang: But This bright color to a dark color. You just feel like, you know, you're, you it's you know, just feel calmer, really. You have to try it yourself and we can get that wired up. So yeah. I mean, we we we did play around with different, you know, these different kinds of color colors. There's more, that we're not showing here. Some like a map or a home if if, you know, that's That that's you know, the we're so, yeah, we're we're we're open.
George Wang: We're open. But right now, it's the the blue that we're converging on. So I know I went through that super fast. Any thoughts on the visual direction?
David Kawata: I think there's just one point. The product itself, as far as the chat experience, just wanted to highlight one of the requirements was the concept of a white label. Right? So create a clean palette
George Wang: Mhmm.
David Kawata: that can be used into different branding experiences as an embedded. So as we look at colors, the, whether it be more of a white palette that we put color on top. So it's That it can go into, more diverse, embedded experiences. It's just a concept that I wanna be mindful of. But, overall, like, some of the new illustrations, and some of the things, one of the things that I've always appreciated is your iteration times, and I know that we've had really short time.
David Kawata: So, yeah, moving forward with the evolution. So okay.
George Wang: Hear you. The white label comment. Yes. Remember from the,
Sydney Bocik: Can I ask a question for my understanding on the white label just so I understand? Because it the is it it it makes sense to have color in the prototype to highlight the things that will need to be white labelable. Is that kind of the thought process there as well? Just so we understand. Like, because, like, if there is color or logos or or names. Is that Roam, like, a technical and implementation perspective in terms of, like, de ning requirements, is that because then that will help us de ne, like, what needs what should exist as, like, primary colors, secondary colors, where the logos exist.
Sydney Bocik: Does that make sense? What the Yeah.
George Wang: Okay. I guess it depends on so so far what I understand about white label is there's the there's the there's the refinance path and then there's the purchase path. Right? And those are the two directions. Maybe there's others and you're thinking about this as the heart of all these different products, and not a it's so it's in the back end. It's the back end of these different it's not the back end, but it's like it's a combination. It's the conversational interface in Oh, maybe share the same database and and agents and AI.
George Wang: Right? This is a way to bring it to life and, like because right. If, like, we don't show this, like, we'll just be showing you, like, wireframes, like, white and black, and you have to brand it anyway. So this is an example of how it could be branded, whether it is for refinance or purchase or personal CFO. And then, you know, we can take this and, you know, apply a different color, different, you know, sort of, positioning.
George Wang: But but but this the colors and the individuals give it a make makes it tangible for us to give us something to react to. Right? You can imagine this in green or Know, but slightly different background.
Sydney Bocik: Okay. No. That makes sense. I I like the way you, especially when it was where the the explanation cards you had to that, like, really felt like it brought about down to earth or put it in layman's terms where the where the one card you had.
George Wang: Oh, this.
Sydney Bocik: Yeah. And, and also too where you said, like, your your mortgage is working harder than it needs to. Like, I thought the explanations in here, even just from, like, a quick read, were were nice and helpful. And the the pill additions on what else they can ask is really helpful. So it's great.
George Wang: And there are so many ways to do these pills. Right? There's the, you know and so that's what we're we're, you know, we're exploring here, but we're also exploring the portal type. But, yeah, wanted to show you guys this just so you get the, you know, see the breadth of the work and exploration that we're doing.
Sydney Bocik: The the question that I had, and maybe I missed it in the quick overview, is in the prototype tab you showed us the other day, you had, the little collector dots in there. Did we Because it seems like a we lost a little bit of, like, the the progress intelligence in here or,
George Wang: Yeah.
Sydney Bocik: like, clarity or hints in here.
George Wang: Mhmm.
Sydney Bocik: It was there what was the decision to kind of move away from that?
George Wang: Oh, it's still there. It's still there. It's just missed in this, and we're we're just experimenting with different,
Sydney Bocik: Okay.
George Wang: like you said, at the bottom, is it at the top? Like, do we But I know that's
Sydney Bocik: Got it.
George Wang: something that, you know, everyone really liked. So let me just turn it over to the to the, yeah, to to this prototype now.
Sydney Bocik: Bye.
George Wang: Okay. So mobile view, desktop.
David Kawata: The the font's really small still in the in the in the and I don't know if this is as far as the spacing,
George Wang: Yeah. I think
David Kawata: I mean, to understand was that on purpose because there's a lot of, like, gaps.
George Wang: the this the normal screen's a lot it's smaller, so you can you hold it, you know, up to your face. I'm not sure how you how how you guys are looking at this in the browser. But, when you have it in the phone, it's, you know, a little bit easier to read. This is as much I can as I can minimize it. So But, yeah, we're checking everything against the accessibility standards and, you know, in terms of, contrast and using specific fonts.
George Wang: So that I mean, that's on top of mind. You know, we need to create this for 40, 50 year olds. Home address, say, k. That's a page. K. Alright. So it's looking it up. Like, what exactly so in terms of I mean, there's still still a lot of work to be done in, exactly how the agents navigates through everything.
George Wang: We've been mostly focused on, just getting that 80% right? Like, it it's not entirely off course, but to hone to get to that, you know, last 20%, that still takes quite a bit of time. Right? So, we have these tails at the bottom. That's, you know, adopted based on what the agent is saying. So, and and these as you can see, it kind of, gets people to voice their motivations.
George Wang: And they're not hardcoded. It's just we're Hey. It's you know, try to elicit that from the users. And it's not something that you're asking, but it's, like, something that you know?
David Kawata: In the last version, when we had a text with prompts, right? So one of the things that I kind of liked about that, it was still a picker adventure, but yet you could take a path that has been defined. Was that, what's your thoughts about I know I'm not sure yet is a path that can be exploration or can be explored. But, Case they wanted to go completely, somewhere different, now we have a couple different steps to get to that that experience.
David Kawata: What's your thoughts about having the text and the pills so the customer can engage exactly where they need to as quickly as possible?
George Wang: Be able to do the text and
David Kawata: So you you took out a YouTube so yeah.
George Wang: the text.
David Kawata: So, like, I don't see a text option, right, besides the pills.
George Wang: What's on your mind?
David Kawata: Right? So in the last examples You you had the text and then Vineet recommended pills and now I see pills, but I don't see the text. So I don't have a an ability for the consumer
George Wang: Oh.
David Kawata: to at one point to respond to get exactly what they want. So we have suggestions. I'm sorry. I can't see the tech if there's a text box underneath, I'm not sure. Is that what that is?
George Wang: That's weird. Yeah.
David Kawata: I just can't see it.
George Wang: Maybe it's because the way that that the screen
<v David Kawata>Okay.
George Wang: is framed.
David Kawata: Okay. Then now I see it. Okay. Thank you. I'm sorry.
George Wang: Okay.
David Kawata: I just couldn't.
George Wang: Alright. Got okay. How do we lose that? Alright. So is that so, like yeah. Like, I we just always wanna give them you know, ease them in. Right? They might not necessarily know What yeah. Where they where they wanna go, and they just might wanna just learn about refinancing and what this tool can do. So, let's just say I'm not sure yet.
George Wang: Okay. I mean yeah. It's just like, hey. Here's different paths we found for you so you can have something to react to if you're not sure yet.
David Kawata: Can we can you show me that on the can you show me
George Wang: So, yeah, the left side.
David Kawata: on the mobile? Because I like how that expanded on desktop.
George Wang: Mhmm.
David Kawata: Can you show me how that would
George Wang: Yeah. It'd be like this. Here's the three scenarios. Click on this scenario. It pops up and then it's like, it's it's just one sentence. Otherwise, if you, you know, populate it with everything, it's like, well, it's a lot of data, a lot of information. Right? And as you can see in the design, we're playing around with, like, generating an image. And maybe that'll make this feel more spacious, you know, breathing Roam. You know, maybe there's, like, a metaphorical image we can generate of a leaf, whatever symbolizes that at the fast track.
George Wang: Right? So it gives a little bit more room to, to, you know, to use the user in. But right now, the breathing room, what changes, your monthly payment, some charts, when it pays for itself. And the bigger picture. Now this is this, these content pieces are you know, this is where we're looking for your more of your expertise. What should be the firstpiece, the second piece, the third?
George Wang: What what like, what should these be? Right? We have we have these, the ability to adjust, Does the target does that adjust this recommended, scenario? Or does that just, you know, keep it down here and you can play around with it? And what we also built is the, you know, if the like, they can just click a button and simplify and switch across these different Yeah.
David Kawata: I like the I like the expandability. I like the simpli cation. Right? So, again, try to give the recommendations, but still enable the consumer to get how they want the information. So I like that that change from just previous iterations. So thank you.
George Wang: And and you can hey. You can it's not working as long as but clicking the save button, you know oh, hey. Pedro.
Pedro Vieira: No, George. You may nish the the safe button. I can make the question after.
George Wang: Okay. Yeah. The save button will bring it into the, the clarity. The clarity report. And so what we moved the the dots up here. So as you're saying, you know, I have six point ve percent rate. Oh, that didn't work, but, I don't know. What's what's my breakeven?
George Wang: Okay. Well, it's a prototype. So, what happens is the the dots will let's just see. Roam either six. Right. See? It just went on top, and we're we'll we'll adjust this so that it it it's more cinematic.
David Kawata: I had a question regarding the columns of text where you're cutting o the the ows. Like, we seem to have a lot of real estate from, edge to edge, but and, you know, the conversation is going left to right on the the alignment. Is I know it's stylistic, but, it's it seems like we are yeah. There there seems to be I don't know what the right Design standards are 20 sixes, but, it seems like we we are leaving okay.
David Kawata: So, yeah, still, we're leaving a lot of room that go reading down versus continually going straight. As for ease of ability, tell tell me why.
George Wang: Okay. I think the maybe the reason is as I'm changing the screen size, it's screen sharing a little bit differently. So why don't I just share the entire screen? Because last time, what you saw is a it's like a it's a fake iPhone inside a browser. And this time, we made it a like a browser that you can go yeah.
David Kawata: Responsive. Yeah. Yeah.
George Wang: It's responsive. So and I'm just sharing that window. So as I'm adjusting that window, maybe I don't know what you guys are seeing. But What? Oh, shoot. I just,
<v David Kawata>I I do see in in,
George Wang: trying to reshare again.
David Kawata: like, the Apple text. Right? They they don't oh, they go, like, three quarters a line for both sides, and then that will allow you to have the sidecar, you know, illustrations, pops, highlights, which is cool. It just seemed like it was really kinda cut o in that.
George Wang: This is this is what I see. You see the window? No?
Lucas Menegazzo: Yeah.
George Wang: K. Alright.
David Kawata: Nope. And I understand the iterations are quick,
George Wang: Yep.
David Kawata: and you actually done a really nice job when we call it hours and responses on some of our notes. So, again, I'll take I'll take that all with a grain of salt. But I do see that we have a pretty 5050% column split on the text versus the three quarters, which, just seems to be more horizontal reading versus linear. Or and I don't I don't know the read readability, the science on that, but doesn't seem to be best practices. Just a little yes.
George Wang: Are you are you referring to the width?
David Kawata: So, like, that that box yeah. Yeah. So yes. The width of yeah.
George Wang: Oh, yeah.
David Kawata: So it seems like fty fty
George Wang: Right. Yeah.
David Kawata: columns versus the three quarters or 60%. It just seems like I have to read more, good lines versus cross. And I think really.
George Wang: Yeah. That that mean, that that's part of the, just the, I guess, the the trade offs, the sound size of just having to move so fast because, you know, I had on the on the design, like We're indicating the visuals and in the prototype, we're trying to play around with the the interruptions, you know,
David Kawata: Yeah.
George Wang: how these safe how safe works.
David Kawata: So like in,
George Wang: And so yeah.
David Kawata: in, we're cool. So like in screen, drop two, three, like On-It that one, but it is now further along. So again, for readability again, I would just wanted to ask if that was intentional or if that was again to what your point is. We're just trying to build fast and get more nodes and concepts out. This sounds like it is more conceptuals trying to get out, make sure we have the the everything directionally correct. Versus a stylistic, choice. Is is that what I heard?
George Wang: The Sigma is more stylistic.
David Kawata: K. Thank you.
George Wang: Visual style and in terms of functionality is not the most updated. And on the prototype, it's more the functionality. You click this and that gets you that gets you where. And then we're we're in the process of marrying these two things together, and by the end of the next week, we'll have a design system.
Pedro Vieira: Good. George,
Charlie Gann: I think oh, sure.
Pedro Vieira: I'm sorry,
Charlie Gann: Go ahead.
Pedro Vieira: Charlie. Please.
Charlie Gann: No.
Pedro Vieira: No.
Charlie Gann: No. You go ahead. You had your hand raised.
Pedro Vieira: Okay. Very quickly. You mentioned, Ampa, that, we should have some sort of imaging, right, on the the application, like, that one in the brie ng room. Is your expectation for that to be dynamically generated or customized to, like, each user or for us to have a set of static assets that we, pop up in a certain condition?
George Wang: I mean, I think this is what the long term conversation of, you know, how static versus dynamic are these scenarios. And there and there's, you know, there's deep dives I think we should do maybe, you know, today to think about what what are the sections in the in the in the scenarios. So far, it's just, you know, us reading whatever we can in the mortgage space. But Maybe that's where we can use more of your help. David is, like, the firstsection, it it should be this.
George Wang: It should be about, it should be about, like, your current state. And then the section is, like, the you know, where, you know, how how this scenario compares to your current state. And so, like, the sequence of things, it's it's not something we have spent a tremendous amount of time on yet. That's that yeah. That's something we love to Mhmm.
David Kawata: Let me reword the question because there was a question that I also have. Right? So AI has been amazing. Every week, there's and and Claude and illustrations are becoming even better and better. Right? So the I think one of the questions is as far as the whatever the date is today, as of today's date, do we think that some of these illustrations as we create these prompts to create it, do we wanna create a library of static The calls to to create these illustrations like the breathing room where we know it's consistent for that topic, or do we create them generatively on the AI basis for the same talk and prompt and generated Roam your recommendations.
David Kawata: That that's kind of what I'm interpreting the unknown right now. Do we build libraries of calculators? Do we build libraries of, illustrations videos? Or do we create prompts that that could grow in the dynamic basis based upon technology, speed, time, and, you know, consistent outcome? What's your thoughts?
George Wang: Yeah. I mean, definitely the prompt is it gives us more more learnings. Right? It's it allows all of us as we're building this company to learn whether it works or not work. And it's gonna run more risks, and we're gonna learn more. It's the process of building a start up especially in, you know, using technology that's constantly changing and, you know, vice versa with the images. And so it's it's more secure. Know exactly what's gonna be shown? So that it's, that's, that's the choice.
George Wang: I think, you know, like we can generate these images, but then you you're gonna have to add a lot. There's gonna be a lot more complexity involved if you want them to stay, you know, within the guardrails.
David Kawata: Yeah. So I think, like, these type of, illustrations, having a library for talking points so that we create a the visual consistency is great. The, so I think that would be a ne starting point. And then we could figure out the dynamic natures and consistency, but like on calculators, like I was really impressed with the uidity of graphs and calculators and the and how it could be responsive. What is your thoughts on that type of interaction?
David Kawata: Creating those guardrails. Do you think we build these componentize and then hyper personalize the data into the the components, or do you believe that we create prompts into creating the graphs and the reports and the calculators, and we just evolve and deal with, you know, stubbing our toes and just deal with it. What's your general thesis on that?
George Wang: So right now it's using JSON render. So what JSON render is, it's a generative UI interface. So we're we're the the agent is creating these charts in real time. We didn't create this like, we gave it a chart box. It's like, here's, here's a line chart. Yeah. And here are the variables that the line chart would need. And the agent is generating those variables, passing it to the line chart and say, Hey, like it makes sense in this place to use a line chart and here's the variables. And so it's generating Gonna look like, but, you know, here here's you give me a a a template.
George Wang: And so I'm gonna use that template. I'm gonna serve the illustration or, you know, in this case, graphs to you. And all all the all the interactions is precoded, but but the lines and the math that is, you know, added to the template, if that makes sense. So there's several different there's a calculator component. There is a, you know, line chart. Bar chart, and we can create as many as we want. But I think, ultimately, what is that? What is that?
George Wang: And and we could maybe just give a few pass for the agent. You know? It's like, what do you think? What agent, what do you think? Like, what depending on this person, you know, like, like, what what what what instructions do we provide the agent for them to make it make it the good make the decision that we would make? There are certain kind of person, like, here's what you show first, here's what you show second and third.
David Kawata: And that's what okay.
George Wang: Right now, for the sake of experimentation, it's just, like, letting an agent run that.
David Kawata: And that's what the exciting part about AI for me for the visuals and personalizations at this point. Right? So if, I'm dealing with the 25 year old versus a 50 year old or 80 year old, right, the text or the different spacings could be more dynamic, not necessary in the MVP. Right? But those graphs in that we will see that the time to different ways to iterate, and display that. That that is exciting to me. So alright. Thank you.
George Wang: Mhmm. Yeah. Good to know. And just show one last thing. I think this is hopefully, you can still, extend it, the desktop view. So as, you know, and and, like, keep remembering this metaphor of you're with a loan officer or you're not with a well, you're with somebody that's helping you, that's helping you understand, you know, spreading information on the table so you you really have a chance to absorb everything.
George Wang: In mobile view, we've, you know, we in mobile view, you know, we have these pinned notes. It was at the bottom before but, you know, we just moved it to the top. And okay. That's not working either, but, something about prototyping. It's just say, show me my Clarity report. And okay.
George Wang: No promises. Okay. Well, it's I don't think it's Working right now. But, what we had before maybe I can't even pull up a screenshot.
David Kawata: What was your thought on the blue dot? I I don't what I I I understand it may be used for thinking or some type of animation. I right? I'm assuming that that blue dot, You brought it up, but I it it seems a little out of place.
George Wang: Yeah. It's it's like a right. It's like a it's an avatar. It's it's it's like it's a character. And it could be it could be anything. For now, we just, you know, use this three- dimensional cool looking AI that is going to replace what you see here because we don't have the, you know, other all the, We don't have the image here, but in the other design le, it's all it's like a three-dimensional.
George Wang: It just makes the makes the whole experience pop a little bit more. It makes it feel alive. If there is another character for the Roam advisor you have in mind, certainly welcome, you know, welcome that. But this is what it would look like if it was working on a desktop view. So as you Yeah. This scenario is what basically is this where you can adjust the scenarios, look at the scenarios, and then whatever you save will also appear on the left side.
George Wang: So the, the clarity report, the scenarios deep dive view, and the conversations are three pieces that can work together. We're a little bit past time. Anything else we should talk about now? But regardless, I think we should definitely get some more time to, to just, you know, yeah, work together.
George Wang: Sorry. I'm kind of it's been a long day.
David Kawata: Enough. Why don't we schedule that extra time, when you you think you can call it Monday or what have you, when you want to add those additional working times? I think there was a lot that you guys did within the the time Roam. So I wanna recognize that. When would these, Figmas or these images be available for the team to review?
George Wang: They are in the board now, so the team can review and make comments. And we're we're in the process of, you know, taking the say taking the Figma screens and really merge it. So take everything with a grain of salt. I mean, there's maybe things that you like from the Figma, from the prototype. We won't show that as well. So everything is, is happening in real time. But, you know, any any like, everything that that we can put out there ffi for you guys to react to is information. We can then use to, you know, evolve evolve the the the product.
David Kawata: Great. Again, like I said, I think the evolution on a day to day basis has been jumps. Right? We're doing jumps, and so I think that's great. I think one of the feedbacks that we have from the team was to try to understand some of the logic on pathways of the decision you process. Is there a way that you can maybe have your team maybe, you know, reveal the Komodo a little bit in the sense of how some of the decisioning process that you got tofficertain process that would help us, I think.
David Kawata: Moments faster so we could support versus, like, having that con ict of, like, why did we get here? So if you have your design team,
George Wang: Ja.
David Kawata: it's maybe expose that if possible. I I know that the the team has, like, well, how do we do this and how do you you made that, font change. Right? That was in response, I think, to the fonts. Right? And you you you settled on this new font. And so any of those type of If you did, that would help us journey faster with you versus always not not always, but making it feel like we're challenging. I think we wanna swim with you faster if we could have the insights.
David Kawata: I think that would be helpful too. Okay. Cool. Charlie, Vinny, I wanna get your inputs before we jump real quick.
Charlie Gann: No. I'm sorry. I joined a little late. So I'm still kinda processing it. I'm just echoing what David said that's been you know, we've made a lot of really swift directions in the design change, which I think have been wonderful to see you guys iterating so quickly and and making so many changes. But, yeah, just to reiterate what Dave was saying, like, understanding the thinking behind it. You know, what what's our approach here? And, are we are we are you know? Design design logic behind, the choices, the font size, the the colors, any type of elements or components that we're we're building, I I think would be helpful.
Charlie Gann: Just so I know from from my perspective, like, once our, time is up, you know, taking it and running with it, and building on it and growing it, like, understanding the foundation better would would be really helpful.
George Wang: Yeah. Yep. Next week, by the end of the week, we're, building a a design, starter kit, you know, as as a system. Yeah. And, we can't say it's gonna be everything, but it's gonna be, like, you know, the key the core the core components, and the interactions that, you know, with these with the rationale altogether.
David Kawata: We can get the rationale before the deliverable. That way, that would be fantastic, though. I appreciate that. And then, Vinny, do you have the last minute last word here,
George Wang: Mhmm.
David Kawata: sir?
Vinicius Ferreira: Okay. I just want to say that, George, you being very responsive, to the feedbacks. I just want to say that, I'll be providing my feedback on Figma. So yeah.
George Wang: Appreciate it, Vinny. No. Your feedback has been really, really good as well as Charlie. So we really appreciate you guys and David and everybody.
David Kawata: So, again, again, I do appreciate the cycles that we're going through. I know it's, a little bit, we bump our toes a little bit, but we're we're we're getting through that. So thank you for for that. George, last word?
George Wang: Oh, it's been really fun, and, we're gonna we're gonna keep going. I'm excited to to show you guys or to On-It show you guys, but, you know, hear your feedback, and then we'll work together and get to that, you know, get to that get to the goal next week.
David Kawata: That what that means that next week, because we're kinda getting to the nish line, what are the dates that we have? Do we only have two, Sydney? Do we only have two calls scheduled, or do we have to Georgia point, maybe there's another working session that we should throw in there before Here's here's the package.
George Wang: Well, I think another word concession will be great. So next week is the third week. This is the second week.
David Kawata: Okay.
George Wang: Well, we are, we're, we're moving very fast. Next week is the third week. And then we have a one week after that is the finalweek. So by the end of next week,
Sydney Bocik: There are recurring invites already, I believe, on the calendar for next week, On Wednesday and Fridays, George. So, I know the firstweek was different. This week was different. So if you want changes to the Wednesday and Friday existing meetings, let me know. They're already up as placeholders. Or if you just wanna add an additional Just let me know.
George Wang: Okay. Let's, we'll coordinate with you, before I know that.
David Kawata: Alright.
George Wang: Okay?
David Kawata: Perfect. Appreciate you guys.
Sydney Bocik: Thanks,
George Wang: Alright.
Sydney Bocik: guys.
David Kawata: Thank you.
George Wang: Take care.
David Kawata: Pedro, you're in Georgia today?
Pedro Vieira: Yes. We'll we'll be for a while.
David Kawata: For a while.
Pedro Vieira: Yeah.
David Kawata: The July. Right?
Pedro Vieira: To July. Yeah.
David Kawata: Awesome.
Pedro Vieira: To July.
George Wang: In Georgia.
Pedro Vieira: Then Poland.
George Wang: Not Poland.
Pedro Vieira: Yeah. Yeah. No.
David Kawata: And George
Pedro Vieira: In Georgia now, Poland in July.
David Kawata: and George are you?
George Wang: Mariana, I don't believe we've met. Sorry, David.
David Kawata: Go ahead, please.
George Wang: Mariana. Hi. Okay. Alright. Good afternoon. Hey, Sydney.
Sydney Bocik: Hey. Good to see you.
George Wang: David, were you gonna say something?
David Kawata: I was just asking, are you from Canada, or do you Consider it Roam home?
George Wang: I'm in a lot of places. I mean, I'm in San Francisco right now, actually.
David Kawata: Oh, okay.
George Wang: Yeah. But Toronto, I've been for the last few years where I grew up. China is where I was born in. The US for the last ve years, mine's the last two, so it's back and forth. You know?
David Kawata: Very cool. What part of China?
George Wang: Around Beijing,
<v David Kawata>Okay.
George Wang: the port city called Tianjin.
David Kawata: Mhmm. Yeah. Major manufacture.
George Wang: You familiar with it?
David Kawata: Yeah. Major manufacturing hub. Yeah.
George Wang: Yeah. They make airbuses and a lot of drugs. A lot of, the COVID testing kits were imported from there.
David Kawata: Cool. Alright. Let's get let's get to it.
George Wang: Alright. Yeah. I'll share my screen. That's good. Wait a second. We got the Sigma. Alright. So before we so let me just go to the Roam first. Alright. So we're in week three. Week three, Model response behavior, develop wireframes.
George Wang: So we we, you know, started on both.
David Kawata: Don't I don't see anything.
George Wang: And
David Kawata: Are you showing your screen? Thank you.
George Wang: is it, is this can you see it now? Okay. Alright. Yeah. So the conversational model and response behavior, that is a, you know, ongoing project, and we have formed a new mental model of how And we'll show you that. And, also, the goals for this week is to establish a design system foundation and to to nd the key interaction moments that that really, you know, nail down what the, the magic moment, would feel like.
George Wang: And by the end of this week, we have a decision gate, where we're gonna show a, you know, the full, product end to end so that we can align with bits. And next week, we're going into, developing that handoffpackage based on our alignment session. K. So in terms of this, model that we have developed, the question that we keep coming back throughout, you know, thinking through the logic of the system, How do we continue the experience, after we show the scenarios?
<v George Wang>Because right now, you know, you enter your information, the systems ask you for your more information, and they eventually show you some of the scenarios. What if you don't like the scenarios? And and that's just something that we felt like, you know, is something that, you know, we need to think through. And and so in terms of how do we do that, we basically, broke it down into, user action AI advisor action.
George Wang: So user would initiate something or the system would initiate something. And then how we're thinking about this interaction is Conversation and then there is the shared artifacts. The shared artifacts, it could be the clarity reports, it could be the scenarios, and then all of these things are working in conjunction in order for the system to develop a mental model of how the user thinks, what their trade offs are and in order to get to that clarity, the definition of clarity is for the user to feel like they know what to do and they might not or they might not be.
George Wang: Right? It's we're giving people the ability to, you know, offerambiance On-It. And but it's like this continuum where in the beginning, they don't know what their options are and they don't know, you know, like, how their trade offs how they would trade off, like, monthly payment versus term length. And they don't know that. They don't even know these concepts are connected, right, until they come to the experience. And so how like, what is the best way for the users to, you know, Get a understanding of these concepts and internalize it.
George Wang: And just giving, you know, whether it's drops or magic moments, like, simplifying the language is one way to do it, but we're thinking how do we go beyond that? How do we create an experience where they can interact with the system and then watch how the system responds and the clear and, you know, the here's the scenarios that the system develops. Now you give feedback to scenarios and the system learns more about you based on your feedback. This loop is how we're, you know, thinking about this, this interaction.
George Wang: And it's not just relying on simplify simplifying information or analogies, but it's, like, really getting people to experience, how the scenarios will evolve if depending on the the the, you know, the the the different ways of making the decision, the trade offs that they can, you know, test out. So we're calling this the collaborative sense making loop. At this adviser action where, the, the adviser in the conversation part of the the experience like, the goal of the adviser is to Roam for insights Insights about the user, and that could be mean a lot of things.
George Wang: Right? Like, motivations, constraints, preferences, how much money they're willing to pay, maximum a month, and what's the minimum. And so there's it's it's a continuous process of getting information and we don't necessarily have to wait until all that information is. In order to create the scenarios, but, like, as more information is gathered, the kind of artifacts will be improved. So do a prompt will prompt the user to know, for ask you know, it'll come up with the questions on its own based on what where's where we're try getting trying to get the system to go.
George Wang: And the user would share, you know, facts, goals. Here's my ad from here's my address to you know, I would prefer to have a lower monthly payment than no longer. Yeah. Because those two start those two things are opposing. And and refinance, in general, it's like the eight dimensions. We're just thinking through that. There's eight eight seven, eight dimensions. Some are, you know, the the primary ones, other ones are, you know, being, as a consequence of making some it's it's just the the the number of connections between these ideas is so many and the goal of the system is really trying to Fifigure out what is not applicable, what is not relevant, and make the make the problem space as small as possible to t the, you know, the the needs of the user.
George Wang: And so that way, it's this kind of back and forth between the user and and the system. And so the user would share facts and then that would be shown the Clarity card. When the the user has the ability to see in real time as they're sharing information, getting, you know, the system understanding it and putting in the Clarity card. So so it doesn't feel like, you know, it's it's it it doesn't feel like there's a continuous it's just it's just more text and more, you know, more messages sent back and forth, but it's all accumulating there.
George Wang: And by looking at the credit the the credit card, maybe the user will, you know, change their mind about, you know, their their, like, what maybe they after, like, experimenting, they realize, hey. You know what? I actually like, I'm okay with having a longer term for not as much of a, you know, reduction in my monthly payments. So they have the immediate ability to go back and enter the Clarity card, and that'll prompt a conversate prompt, the AI to to then ask for context.
George Wang: And then they'll share more context, and so there's this feedback loop going back and forth between the user and the system. And it's just like in a real advisory interaction when you're talking with a friend that's here to help you. Options. There's the conversation that we're having, and then there's the artifact that's on the table. Like, I would write something down, and I would show you just, you know, just like a, you know, a real in person interaction. And and they're they're, like, they're these two layers channels are interconnected.
George Wang: So when you change something, you know, when when you make a comment about our artifact where we need, like, say, I reject I like this scenario. I reject the scenario. I I edit the scenario based on the sliders. It'll trigger the agent. Ask for context. And so it's like it's it's getting a ever more granular view of how the user makes decisions, not just what they say, but what they do as well once they see something. That make sense? Of course, there's, you know this is a very, very simpli ed model, and there's a more Know, like, there's a longer model and then there's even a bigger model where we think about, okay, what do we show in the Clari report?
George Wang: How do we divide it up? And, you know, like, all the scenarios do we have enough information to share to to show the scenarios? You know, what does the comparison logic look like and what are the types of actions the user can take and how does that feed into the the the system's understanding about this user and, you know, what need what what part of that understanding is shown, not shown though. So how does it Con icting information. Like, I wanna lower monthly, and I also wanna lower term because the user doesn't know that these two things are opposed to each other, right, in most cases.
George Wang: And then, like and then, you know, how how do we, identify these contradictions and then help give them a frame of reference so that, and and yeah. And and and kind of Drive towards, like, converge towards more clarity around the sort of around exactly if you if you're faced with this trade off, like, what would you do?
George Wang: So, like, in in terms of developing that mental model, the the the the more the systems develop it, the more the users develop that clarity for themselves. And That's that's the goal. Right? It's just, like, what I wanna do or which direction I wanna go in. It's not the exact thing. And but more importantly, why? All the whys, all the trade offs that you've captured for me. Before I move on, do you guys wanna talk about this part? You know, do you have any questions, thoughts?
George Wang: I keep going there. Yeah. So, you know, with the visual system, there's really there's really three decisions that we wanna make. The typography well, actually, there's really, yeah, three decisions, typography, palette, and the clarity system. So we talked about the, you know, we showed the the the dots. It's you know, we will get to that we'll get to that in the end. I sent a message after the, the meeting last Friday and, have Sort of the the the interaction that I wanted to show you guys and and decide whether we're gonna go with the, you know, the big dot direction or just the small small dots.
George Wang: So in terms of and about regarding the comments about, showing enough rationale to show, like, how we moved from you know, moved between versions, this is this is something that we thought was quite important. So these dots were at the bottom before. The challenge with that is when your thumb is typing, it might occlude the dots and you might accidentally hit the dots and what we how we we move that to the top because the top is, like, the orientation that's telling you where you are, the the context that you're in, and the bottom is where you would, you know, that's the action zone.
George Wang: This is the reading zone and that's the action zone. It's a little bit more organized. So that that's one of the the key changes we we did for, in, you know, in in this version. So you mean the last version that I want to to to bring back and highlight. In terms of typography, I mean, one big decision that we wanna, you know, decide right now is just do we wanna go to a sans serif, heading or serif heading. Both versions will use inclusive sans as the As the, the body.
George Wang: So heading wise Mhmm. What do we feel? What do we feel like? What, you know, best, represents the brand? Pedro.
Pedro Vieira: And, George, just a second.
George Wang: Yeah.
Pedro Vieira: If you could hide the Figma UI, I think that's gonna, allow you to to enlarge the
George Wang: Okay.
Pedro Vieira: picture a little more. Next.
Charlie Gann: George, just so sorry, David. Just so I'm clear, is the question between San Sarif and Sarif or between this San Sarif,
George Wang: Yeah.
Charlie Gann: this Sarif? Like, this specific one or just in general?
George Wang: Well, I'd love to hear your comments on both. The what we're using right now is, Roam, Crimson Pro for the serif, and for the sans serif is the, it's rethink. We think sans.
Charlie Gann: Yeah. I have, I mean, I have I have a few thoughts on it. For so from a SERIF standpoint, SERIF adds a lot of personality. You get a lot of, you know, with with the ligatures and the different there's just more personality that can come through,
George Wang: Yeah.
Charlie Gann: but rarely reads as modern or, as tech oriented, I think is where we're wanting to take this brand. I guess it also sort of depends on I think about this more as a product, in a digital product versus our marketing and and speaking top of funnel. It's a little bit different. I think in this application personally, and I'm open to other people's opinions, I I lean towards a a sans serif, for really just I I think there's gonna be a little bit more of a sleekness, I think Roam a UI standpoint as you talk about The conversational elements and the different components on the page, I'm I'm leaning more towards San Serif.
George Wang: David or Cindy or anybody else, you guys, have a point of view on this?
David Kawata: Vinny, I'd like to hear from Vinny.
Vinicius Ferreira: Oh, on on my side, I would like to see, I have it's it's hard for me to to determine, like, right now which one is better. I would like to see, I don't know, like, another, piece of, like, illustration. I I I mean, I would like to see something more, like, as Product. I don't know. Sorry. To to understand which one, like, better, ts the the picture of this.
Vinicius Ferreira: Because we are right now, we are just looking at app. But I will I will I think this is more aligned to the brand. On my side, I it's a kind of preference. I prefer the serif one. I think it it just bring, like, people. Like, nuances or but it's just a preference. Right now, I will I will like to see regard to this to decide, like, another piece of material in which we can see, like, a composition between, like, I don't like, maybe a landing page or something like this.
Vinicius Ferreira: It's just something from I'm out.
<v Charlie Gann>Yeah. I I agree with you. Now I think that's where I was going with, top of the funnel versus with within the product. It is hard to tell. Like, in this image by itself, it's it look feels more like an ad, and the the one on the right, the serif is is, a really lovely application there. But when you get down into the actual product, and I think about as I'm, you know, three conversations deep.
Charlie Gann: It it it should take a little bit of a step back. Right? Because it's more about the conversation content,
George Wang: Yeah.
Charlie Gann: versus whereas this feels very much top of funnel. You know, there's this lovely picture of
George Wang: Right.
Charlie Gann: a really beautiful home. You know, it's, it's hard to, to judge from, from this out t. I think that's where, where you were going, Vinny.
George Wang: Yeah. That makes sense. You guys will see more Wednesday, but we're leaning towards the sense thereof, you know, for the product, and you'll see in the subsequent pages. And, I mean, that makes a lot of sense that, you know, we can have more, variety. The, you know, top of funnel landing page, the brand, the product. And I agree with you. It's, it's of course, better to have a cleaner cleaner text. Okay. Then maybe we can come back in the in in the end of the meeting to just make a call of whether we can just go ahead with the Sentara.
George Wang: In terms of the Colors. So, you know, this is where trust either happens or it doesn't. Right? So on the left where we started, it's functional, reads like a generic, ntech UI, and there's a lot of cool blue, a lot of Chrome. The whole thing feels like an app interface, but it doesn't really, you know, do much to support the emotional state of someone that's making that financial decision. And on the right side is where we're taking it. We're not, chasing a pre pallet, but it's, actually, more about register.
George Wang: And it's optimizing for three things. One, it's calm confidence. The screen feels steady. It's not loud. It's highly readable on mobile. All the fonts are 14. The body font's 14, labels no. Body font's 16, and the label's 14. So it's, perfect for, you know, aging people that, you know, 40 can see is bad eyesight, especially with the the topography in inclusive inclusiveness. And, yeah, and we're only using the, accents sparingly, you know, when it's turned.
George Wang: So, what you're seeing in this option two, it's, you know, just warm, like, neutral to the base. The product feels human, and not too clinical, not like the the one on the left. And the cool anchor in the middle, the blue one, it kinda keeps people grounded. You know, in nance and trust, and if there's somebody that's, you know, thinking about Financing is a lot of stress, and the blue kind of makes you, calm down just a little bit. And, yeah, the the control of action is only, you know, used for moments like when you're capturing, data from insights, from APIs, or from, somebody telling you in a conversation.
George Wang: So we'll show you some examples. Yeah. So, what we wanted to decide today is basically whether we wanted to feel more cool, cooler like a financial product, quote, unquote, or more warm and, like, an advisor on the right side. Hey, Pedro.
Pedro Vieira: Just a location here, George. Would it make sense for that to to change, to match the emotional state of the user?
George Wang: Perfect.
Pedro Vieira: I mean k.
George Wang: That's a great question. Yeah. I mean, it's hard to understand the emotional state. That takes a lot of data, but in terms of gathering the insights and we have some examples here, you know, the the colors of the, the color of the depending on what is the inside that's being pulled. There's a color coding. There's, probably something, green means something, and then depending on changes too. So it's it's dynamic.
George Wang: So is it, is it safe to decide that we're going with option two,
Pedro Vieira: Nice.
George Wang: or is there anyone that's that prefers option one?
David Kawata: Don't think that's a safe assumption, but I think it's, for our conversation. Thank you.
George Wang: Okay. Alright. And lastly, in the in the, wanted to show you the the the decision about whether we show a big dot, which is not a character, but it's more like a heartbeat and a presence that basically, like, what I was sending in the in the envelope on Friday. I wanted to show you guys this portal type and That's there's a way to restart it.
George Wang: Okay. Alright. So, yeah, this is the beginning. So, you know, conversation. As you're sharing information, the dot, it signals that your you know, the attention of the adviser is on whatever it's next to. Right? It's, like, just like when you're having a real conversation with an advisor, you you can see where they're looking, you can see where they're pointing, and this is, indication of what they're focused on.
George Wang: And Know, that's my name, and then it oats up to the top, and it increments for items clari ed. Because we're thinking if we, you know, keep on incrementing, then adding more dots, then we're gonna run out of space. We're just gonna keep one. And, you know, it's like, I I know it. Then can click on that and go to a different page. And on bottom, there's, there's the the the actual insight that it took out. Bob. Since you bought this property, this is how much equity you built out.
George Wang: And currently, the the color coding, in this is Robert is it's by, not the theme of what they're saying, but rather whether it's a fact or it's a, as a preference or it's a constraint, because, like, you just can't predict, you know, what are the, you know, million one things that somebody would say. Know that the system will nd it relevant as a, as a way to help them converge towards a, you know, care clarity.
George Wang: So if you don't know that, then it wouldn't make sense to specify color, different colors for, like, based on the theme, the the semantic, but rather, like, on a functional sort of category, fact, constraints, preference. So equity is also fact. And I wanna decrease my monthly payment, so that's, a slightly different On-It, color coded gray. But what you're the color coding in this is, the team is playing with that still. So what I'm showing is just the the purely the, the the animation, the the functionality, some monthly max.
George Wang: And then when you go to that page, all the dots ow back down just like how you saw in the prototype that we shared last Wednesday. In terms of the colors we're gonna use, it's gonna be it's gonna be the the palette that you saw in in week one. The yeah. We added we took a maroon. We added green.
George Wang: So that's that's how we're envisioning how the these big dots will work. The alternative is to just Not show, you know, the where the advisers, quote unquote, looking at. Instead, just pop a little dot up to the top when you, when you say something. Is that clear? And what's what what like, what's your, what's your guys' take on whether we keep the big?
George Wang: Big dot or or just use the small dots? You wanna take a do do you wanna you know, can we decide that on this call, or do you guys want to take some time to to think through?
David Kawata: Will it take a moment on that?
George Wang: Yeah. Okay. So that's everything we wanted to share
Charlie Gann: Yeah.
George Wang: this morning. Yeah. So, again, this week is just focused purely on getting to that finaldesign, more or less. And next week, we have one more to, cement the the design system and develop the handheld package. So that's what we're we're driving towards. K.
David Kawata: Is gonna be a Figma or is in Figma now?
George Wang: Yeah. It is in Figma now.
David Kawata: Great. And the interactions, are available as well?
George Wang: The interaction will yeah. We'll we'll put that we'll put that in the Figma,
David Kawata: Got it.
George Wang: the prototype.
David Kawata: And then, so, Charlie and Vinny, if you get a chance to take a look at that today, that'd be great. We can, George, I think when our next call is on Wednesday. Correct?
George Wang: Yeah. Thanks, Carlos. Have a nice day.
Charlie Gann: I'll I'll try to get some feedback in here for you, George, today.
George Wang: Yeah. Sounds good. And let's grab a let's grab some time together, Vinay and Charlie.
Vinicius Ferreira: Yeah. Yeah. Me too. George, something that'll be very nice to have, they're gonna, like, guide us. It'd be, like, to have some kind of, like, concept, like like, a design concept. So some something that we can, use as to evaluate, like, Your, decisions or your team decisions. For example, I don't want clarity, the the design concept means this.
Vinicius Ferreira: So this is gonna be useful for us, like, to to try to under identify like like things to bring closer to that concept.
George Wang: Yeah. Let let's, do yeah. We can we can we can I I noticed this time David has dropped in? Let's grab some time and, talk to the team about that as well. And, yeah. So, yeah, let's let's let's, you know, come back in the afternoon. Maybe just Charlie, myself, and and Vinny, and we can we can hash this out. Okay? Alright. Thank you, guys. Appreciate it. Take care.
George Wang: Morning, all.
David Kawata: Morning. How are you?
George Wang: Good. How are you guys?
David Kawata: Well, thank you.
George Wang: Alright. Christopher here? Okay. Alright. The Roam seems to have a call. K. Share my screen. Okay. So alright. We're currently on middle of week three. We, have put together a prototype that we wanted to show you today and, design system that, that we're gonna share along with, key interaction moments.
George Wang: So, I mean, over the course of the last two weeks and even to this day, we're writing, two two tracks that's Symbiotic. They're, you know, a ecting each other. The the the portal typing to figure out the interactions and, you know, like, we talked about how the AI, work this this advisor, and we communicate through with the AI through a messaging channel on how that a ects the the clarity cards and the scenarios, and we wanted to show you that today. And so that's proceeding as well as this visual track to take some of the feedback that we heard after the previous, call.
George Wang: And Turning into a system that we can then marry towards the end. Alright. So evolution of the three weeks, we heard feedback about, you know, the colors and should be morefineutral. And, so we, you know, went away from using the colors, like, a lot of colors everywhere. We even heard from you, David, earlier, about the the gradients the gradients being used in the home pages.
George Wang: So, we yeah. I mean, so we moved the gradients from the home page, so that it's not the default state of the product. And as you can see here, here's the the new version, but we kept the gradients in the conversation because that's where trust, becomes a guide. And, you know, we use color On-It just it's not decoration. There it's really this ambient mode of communication that then we thought. Homepage so that as somebody's going in to learn about the product, they're seeing snippets of how, you know, they expect to interact with it.
George Wang: The gradient is a way to highlight, you know, you know, certain things that like, facts that we heard about you, we learned about you, you know, opportunities, constraints, and and that and we think that's that's a core element of how the product functions. It's it's it's an advisor that listens to you, that collaborates with you, and helps you. What to do, how to make sense of, you know, this refinance subject.
George Wang: Yeah. So we are, you know, really using ingredients as a way to, you know, communicate. So before I move on to these design principles about how were we using the colors, are we applying it, coming back to the foot typography. And the last time that, we wanted to see more screens and how each type is gonna be used, in those. So we got Crimson Roam, which is a, you know, more elegant, more, character.
George Wang: I think, like, you know, I think, Vinny had said. Right? We we showed you how it looks on the top of funnel, and here are some examples of how it looks in In the actual app. And vice versa, this is a, brie ng sense in the headings. I know that, Charlie, last time you said you preferred more of a, you know, the modern font and Vinny was more on the, you know, the the the serif set of things.
George Wang: Now that you get a chance to see how it plays out in the whole product, is there a direction that we feel, you know, a little bit more comfortable going in?
Charlie Gann: Thank you, firstof all, for putting these together. Actually, really like the balance now, seeing, seeing the serif and how it complements, the sans serif, I think there's something nice to it. I think from the, from last week, I was a little worried that it was gonna nd its way in a lot. But it the way it's been used very minimally, like, kind of the headline of an article or at least, you know Some some heavy amount of content.
Charlie Gann: It's it's a nice balance between, like, you know, this tech techness, but there's still a little bit of a softness that shows some of the humanity, by bringing it in. I just didn't want it to read too much like, you know, like a, like a New York Times article, you know, depending on on how often, how frequently On-It was it it was used. With, with the, with the sheri , then the, the amount of which the balance of which you've utilized it here on the left is, is, is the right amount of balance.
George Wang: Yeah. Very very sparingly and just need, key places, to give it, you know, some, like, character. Okay. So can we are we all any objection as to using The serif fonts, sterily, like, how we are showing it in the left side hand side. Okay. So that's decided.
Charlie Gann: George, before we sorry. Before we we jump, Crimson Roam is is a good, font. It's a good serif, but would it be would you be opposed to me making a couple of recommendations
George Wang: Yep.
Charlie Gann: if I put a few together, and and just because I feel like there's this is a good font, but I feel like there's a couple more that we could take and just push it a little bit, and give it just a a little more of that, calm. Feeling,
George Wang: Yeah. Sure. Hey.
Charlie Gann: that might be really nice.
George Wang: Charlie, really appreciate it. Please, you know, share it, and, we'll we'll take a look.
David Kawata: Phone real quick, Vinny. I just wanna get Vinny's, opinion, Vinny.
Vinicius Ferreira: Oh, no. I would just say that, I agree with Charlie. The idea behind, like, having the serif font is that people have chance to, when we have, like, the serifs, people tend to pay more attention in the information that is there so they can, think more about it. So having this great balance between, like, service and On-It some service. It's it's it's it's good for for users
David Kawata: Okay.
Vinicius Ferreira: to don't understand.
George Wang: Appreciate the the input, Vinny. I'm interested in calling because I I know last time you're already saying the Sarah font.
David Kawata: Great.
George Wang: So now we have, you know, everybody on the same side. Cool. So we're gonna go with the the the one, the Sarah On-It. And, yeah, if you guys wanna send some, alternatives over, we're happy to to check it out and see see what's the, yeah, what's the best one to use. As far as design principles go, you know, this page is really here to make it make the visuals legible so it's not you're not just seeing a bunch of screens.
George Wang: And, yeah, you know, the core idea is really that, you know, most of the interface should leave in a calm, grounded, and just highly readable baseline. And when the product reaches a moment that actually matters, then, you know, that's when we start to bring in, you know, more colors and Face earns the right to become more expensive. So color is just not being used accurately. In this case, it's you know, they're they're carrying different kinds of meaning in the ow, and, and that's how we're intending to use these.
George Wang: So the the, primary, the the, colors are neutral, the secondary is more accents, and then these match moment colors is, you know, just like how you're seeing it here. It, it really bring you know, it's a it's a way to bring you into a slightly different state to accentuate the magic in the magic moments. And we also talked about the, the, you know, the this guide, this heartbeat, and, you know, it's just kind of two things blinged together because there's the thoughts that And then there's the the how we are you, visualizing the presence of the AI next, you know, and how that's sort of, facilitating, if you will, the the creation of these knowledge, thoughts, or pieces and capture these artifacts into the Clarity into the Clarity reports.
George Wang: And, you know, the So, like, we're, you know, playing around with exactly how to execute that because, I mean, it definitely wasn't intending to just be a static ball that moves, you know, left or right and down, but something that's more fluid. It's just hard to show that in Figma. So, I mean, here's a example of, you know, particles and we're, you know, experimented with, you know, using different particles to, like, like, a swoosh effect to take the idea from, Roam know, yeah. That's my name, and then bring it into this, this this tag.
George Wang: Yeah. We, you know, we we experimented a bunch of them and but it's it's it is challenging to execute properly, I mean, to create that kind of effect and, I mean, definitely something that we can continue to explore. But for the MVP, I think, I know you mentioned this too, Charlie, if we don't, you know, the execution is it's it's really challenging. Like, it's not just showing on Figma, but creating it in code. So So, yeah, we're we're gonna let go of this this presence big dot, but the small dots are still gonna remain, and you can continue to explore that and for, you know, post MVP and later iterations.
George Wang: And in terms of how it actually how the system works, so first, you capture the, you know, the name or whatever it is that is important to go in the Clarity report. And as you can see, here's an example of Like, one visual, one is more prototyping. And this is something that I'll I'll show you guys in just a minute, but, it goes into the Clarity reports. And then in the scenarios, we thought that it makes a lot of sense to show the this is the reason why we give you these scenarios.
George Wang: It is because of the things that we heard previously that you told us that are going to scenario. So these three things, again, the conversation, the scenario, and the clarity report, they need to all connect with each other. And so that's why we Oh, you've seen the demo, explored having this, a drawer that, basically overlays the background, which is the context, and you can have a conversation based on what you're looking at. So So then it's just like talking to a real adviser. You're not just talking and then you're you're done talking, you go to a different room, and then you you look at stuff, and then you come back, you talk.
George Wang: Right? So we wanted to create a more uid experience. But, Disregarding the di erences here on the bottom and top, but just to show you how the Clarity dots work, because it's like it's all it's all connected between the Clarity report and the scenario. Okay. Any, any questions, anything we wanna talk about here before we move on? And perhaps we
David Kawata: Yeah. And so, you talked about it's hard for in a Figma to talk about the the expression of the movement, the, right, the feeling. Do you have a interactive, tool that we could play around with so that we could really understand? But that's actually one of the keys that I was really, excited about with Avernco. Right? It's not the font. The fonts are great. Love it. Great. But it's the the movement, the action, the interaction, the transitions. I'm excited about Roam your gaming And and telling that story.
David Kawata: So is there a way that we could get a little bit more feeling of what you are you're proposing?
George Wang: We can let me talk to the team. No. I don't wanna promise anything for the team. They've already kinda overloaded based on, you know, these two parallel ves and try to tie them together. But, I mean, we created, ways to test it internally ourselves. Like Know, come up with 24 variations of how they would, you know, how the the how how the animation might work. And then there's different sliders for how many, you know, particles you want and, you know, how what's the size of this this sphere that it comes together in?
David Kawata: So the the ball the sphere evolving, that that I mean, that's that's, a thing.
George Wang: Yeah.
David Kawata: But I'm what I'm really looking for is the transition, the movement from questions to sliders to clarity report. I'm I I haven't seen the logic of the rule sets to help me right.
George Wang: A logic.
David Kawata: I wanna be able to have a a system that the rest of the team can move forward with. Sense of what you're building that reveals. Like you said, Figma doesn't express it, and I haven't seen any text that express the rule says that you have in mind to tactically imply or to present. So if I could have some text tools, processes, something that could give me a foundation of what you want us to experientially move forward, that would be fantastic,
George Wang: Yeah.
David Kawata: please.
George Wang: Yes.
David Kawata: Thank you.
George Wang: But before then, let me show the, this prototype that we put together, and that's mainly there to communicate the there's different kinds of logic. There's the visual logic and how the transition literally the visuals, and then there's the, the, the the sort of The math, if you will, of how the how the communication works between these different pieces, how information ows across. Let me
David Kawata: As you pull that up, do you think my ask is appropriate for the time frame?
George Wang: show Yeah.
David Kawata: Do you do you so that's part of the magic. Right? The math is one thing. The presentation of the math in a way that creates a reveal or extendability. I'm missing a lot of cars on extendability. It's like, here here's the presentation, and then someone wants to get more understanding visuals from call it tool tips and how that could actually lead into different presentations of the of the drops. That that is And I I know I know to me that I like to get clarity on this week.
George Wang: Okay. Sounds good. We experimented with these different, you know, versions of the drive, and I know that's something that Vinny had, you know, commented. There's a very there's many different prototypes that we build over time, and some of them are more focused on the agents, like, the how the how the agent is guiding the user. Some of them are more focused on the, you know, responsiveness, you know, making the screen larger and smaller, how things change.
George Wang: This one is more built on the realization last week or Earlier Monday. Today's Wednesday. Monday, the, you know, this idea that the there's a conversation and then there's artifacts. Right? The the what what you see and then there's what you're talking about. And then there's the things that's initiated by the agent, then there's things that's initiated by the user, and how all of these these four quadrants would connect with each other. So that's what this particular prototype is about.
George Wang: But, yeah, we'll pull that up. Make that make a note of that. K. So let me go to here. Alright. So listen. This is disregard the the visuals here. I mean, there's different pictures than what you're seeing in the Figma earlier. So, disregard the visuals and focus more on the the Okay.
George Wang: So, how somebody would come in here, they went through their address, looking at the property, and then found the address cards, and we are using these, these, tags, marks, you know, to capture the important pieces from, you know, either APIs or something that they're saying. And so the user has shared their address and one of the payments, and it presents a drop.
George Wang: This is a text drop, of refinancing pay for itself, and, you know, it's just information for you to hey, David.
David Kawata: One of the keys initial start point is the information to con rm who we're speaking to. Right? And then that way, there's a personalization. I I see that's a
George Wang: Right.
David Kawata: but I just see that's a miss from address to, are we talking to George? Are we talking with Sydney? Like, this is what we have publicly.
George Wang: Mhmm.
David Kawata: Like, how do we wanna be called and then hyper personalization on the response. So, just initial thoughts.
George Wang: Yep. Yep. You're absolutely right. I mean, that's part of it. It's, it were yeah. It's just not in this prototype, but as you saw in the in the, the Sigma, On-It was the k Bob because we are able to pull that from their public records. And, you know, we are challenged with trying to bring all these things together into one ow, and that's what we're We're, driving towards.
David Kawata: So on when the on the notes on the card, you put an address and purchase price. But if you if you showed the name of the owners up there, then it's still the same card, same presentation. And then then there's a an ask about personalization. It's like right? So so there's, like, Dick, Richard, like, the variations of how someone wants to be called versus legalism and was a match and one of the the top Them personalized and inspired upfront.
David Kawata: So that is actually a pretty, I think, important date from that experience, but I would defer maybe this is not a full illustration
George Wang: Okay.
David Kawata: of that.
George Wang: Yeah. This is not definitely not. When I get a little bit later, I I think that's what you're gonna see. So these different color codes So okay. Like, I'm I'm gonna be here for another ve to ten years and just capturing that information. And I know that the you know, we gotta play around with the colors here. And, yeah, it's low contrast, but the font is 16 points so that it's you know, everything is very easy to read.
George Wang: So, yeah, it's, just like, like, the the system is asking the user for Know, rationale, like, why, to help me understand to surface to, you know, to kind of turn let's say, to to be able to understand more insights in order to better guide the user. Yeah. Like and then, you know, they're sharing that you wanna do renovation and, and they're agging that, you know, lower payment is Sometimes intention with, taking a money because that would increase payment.
George Wang: So educating the user as they as the conversation proceeds. And they're sharing, hey. This is how much you wanna take out. It's gonna do a little bit of back end mapping calculations, what they can expect. And they're saying that they can't go above $2,000 monthly, and that's a hard line. So, you know, a little little bit more, nothing else with them. And then there's where they see the Clarity report.
George Wang: A lot of the, you know, information they had sent would appear here. So words about.
David Kawata: Real Roam quick. Your your your approximation, your squiggly, it does look like a minus. One of understand Roam math purposes, minus is definitely different. And if our negativity or right? So just just be clear, that squiggly is actually like a minus.
George Wang: Yes. Yes. We noticed that too.
David Kawata: Is that that's actually thank you.
George Wang: Yeah. Yes. Absolutely. This is yes. You're right. Just all these you know, there's a couple different things. On that overall ow and how information ows from one of the you know, into, you know, one of, like, different pieces. And, yeah, something that that we we, we have noticed. So there's the there's the what we know about you, you know, the the facts, what's happening today, how much are you paying, there's the boundaries, like, thinking about what what do you call it?
George Wang: Constraints, boundaries, limits, parameters. You know, ve to ten years, and here's here's how much like, here's it's not a fact, but it's not a, you know, like a desire, preference, you know. And what matters is this is what they're expressing, this is how they're making the decision, and their motivation. And what we found is some of the conclusions, that we're surfacing based on your, and and based on what you're telling us and and opportunities.
George Wang: So let's go continue. So when they this is no. That initially there's there's a couple of different places where the magic moment is gonna come up, and, you know, that being the brief, the magic moment is supposed to wait until somebody's confused and shows up in the scenarios. If we decided to keep that, but also add a magic moment here, that is Like, because when somebody uses this, they don't necessarily expecting that they all will be scenarios that's person like, scenarios that's personalized for them and that this is a very important moment that desires, you know, to stop them.
George Wang: It's like, hey. Look. Here's the things that we put together. I think it's pretty impressive. Maybe it should highlight that. And so we can Look at these and explore different scenarios. So when you're exploring the scenarios, probably gonna show a, you know, truncated version of this, maybe an accordion. It's a lot because it's a lot of information. Right? And you're seeing here's what you told us, and then here is that informs how these scenarios are built. And this this is all built by AI.
George Wang: Reacting to what the information was. So this is the the name that the AI gave it and what, what metrics are important in these these scenarios, and some that's brief description, basics, and sort of warnings or notices, alerts that You know, show which card is in contradiction with, you know, parameters, preferences that they have stated.
George Wang: So they said they this is this is over your $2,000 monthly On-It, and once you increase the term, it's no longer over that. So that disappears, and this gives people the ability to, you know, weigh different trade offs. You change the term in cash out, it's, yeah, it changes the numbers. And this is what it looks like when you're not in violation of these, state stated.
George Wang: So I think a big part of this is somebody go in and understand refinancing. They have to understand these trade offs. And no matter how much you explain the trade offs, playing with it really, you know, makes it come make it helps them internalize these concepts. And Is, say say, my limit is actually 2,500. It's supposed to come up.
George Wang: Oops. I might yeah. I think I clicked on too many things too quickly, then it resetted. But Yeah. Didn't get a chance to talk, but, yeah, as I'm just, you know, pressing accidentally pressing the quick, prompts it's giving me, its responses.
David Kawata: Was that graph a seesaw, meaning there's a direct correlation, or was there multiple adjustments? Right? So more cash out, higher loan Roam loan payment. Right? So it looked like a seesaw, but you have two cars on top. Right? Is that correct on how the interaction was on those two graphs? My point is that that seems like it could have been consolidated with Now go back up, please. Yeah.
David Kawata: So the same variable input, the term, and then but it it looks like that's gonna be consolidated at a one side by side or one one card versus top to bottom because the variables on the term is the same. Right? Now you're just really reflecting the same, dual input from the second card below. But, cool. I I love the interaction. I love the instant response. 88 green and green or red for color blind people. What how how does that play out? Is that just black and gray to a color blind person?
George Wang: We're gonna have to check with the team on that. They were looking through, you know, the accessibility equalizing. I'll check with the team and get back to you.
David Kawata: Right. Because now the red generally indicates bad or negative. Right? So I I don't I I your color choices, but I understand they're comp they're complementary in o ces, the green and red here. But,
George Wang: Yeah.
David Kawata: like to understand the choices you make.
George Wang: Right. Like, the because the different things, the same change could mean People. Right? So,
David Kawata: Right. But colors colors to me is emotionally telling me something.
George Wang: yeah.
David Kawata: Right?
George Wang: Exactly. ffi
David Kawata: And so they're not just, and so the to me, the red is pretty harsh as far as what
George Wang: Yeah.
David Kawata: you and they're generally negative. So just my initial feedback.
George Wang: Yeah. Right.
David Kawata: And also, again, my note on refinances, no one says my target payment is x. They just say lower payment, lower rate. That's a purchase statement that I wanna have a target for a ordability. Just wanna reiterate my note. Okay. Thank you.
George Wang: Okay. Yeah. I mean, we're yes. We're letting users write whatever they want, and we're, you know, as AI is just responding to that. And, yeah, I just wanted to show you that. Hey. Like, I can hey, Charlie.
Charlie Gann: Hey. Just to speak to what David was saying about ADA rules, typically, it's more about contrast than prettier color, and I agree with you. Red is generally used as error or mistake. It kinda evokes a sense of alarm or failure, which I don't know that we want to evoke this kind of emotions.
George Wang: Right.
Charlie Gann: But secondly, it's generally not great for type, especially, like, where you have it on a lighter, like, red on a lighter red background. Maybe Ground would be because it's more about contrast, would be enough to accentuate, hey. This is, you know, some something to pay attention to Roam maybe more of an orange or or yellow as more of an alert versus an error. But, yeah, just wanted to speak to that.
George Wang: Yeah. Yeah. And this prototype is, yeah, just interaction. But, you know, we're always putting prototypes in front of you guys when it's before it's completely done just so we can get this kind of, you know, feedback. And there's a visual track that's looking at what the color should be. But,
Vinicius Ferreira: Hello.
George Wang: yeah. So as you can see, when I tell the AI that I want, a, I wanna increase my my maximum to 2,500 instead of 2,000, it changed, in the Clarity card as well as Me how all these pieces are connected. And if I not sure if this will work, but if I click, you know, not for me, I don't like this, then it'll say, hey. Why?
George Wang: Let's say I want Cash App. Just let's say an argument for that particular it makes it important. So CA just added the cash out to it. So, yeah, whatever you interact with on the screen, the agent understands and can follow-up. And when you're telling the agent something, it, you know, feeds that information to the screen. So all of these pieces are connected.
George Wang: That's the important thing we're trying to apply here. Okay. And then and then and that's all I'm getting you had to hand out, but just very quickly, you can get saved, and then that goes into That goes into your safe scenarios. And, you know, more follow-up, like, when exploring others. So it's the it's this constant interaction between the three pieces. Hey, Vidi.
Vinicius Ferreira: Just want to ask, the, everyone's thoughts about the draw work. Because for me, it it seems that the scenarios are the main focus here, and the conversation, it just becomes, like, the second one. I don't know if this is, it was inversed. I I know. I'm just trying to understand your
George Wang: Yes.
Vinicius Ferreira: thoughts on on that.
George Wang: The thing behind the drawer in the conversation is there is is is that more of a like, how does the system respond to your actions? It's intelligent. It's a you know, we wanna humanize this, but it's it's not human if you're just doing, you know, these two things on different different in different rooms. Right? And, and, there. But, Yeah. What do you what do you yeah.
Charlie Gann: I I think sorry to interrupt.
George Wang: Charlie.
Charlie Gann: I think it's just starting to and, Benny, I'm not I don't wanna put words in your mouth, but for me, it is it's a it was a little jarring when that drawer came up. Like, woah. Okay. Am I starting a second conversation? What's the relationship now? Typically, drawers are reserved for, like, more native behavior of,
George Wang: Mhmm.
Charlie Gann: like, I need to slide up more settings or I need to go in and make change. So it I think for most people on a mobile device, they won't read as this relation this has a relationship to what's behind it. It's starting to now block the information that's important to me. I think if if it's if we're wanting to have a conversation with the content that is on the mobile device, I think the conversation should remain on that screen.
George Wang: Yeah. I mean, there's definitely there there's this trade o between we have a small screen and if they're on mobile and you are you taking up more space if you put a drawer on top? Is it easy to, you know, have to wait because there's a made of phone On-It? There's all these things or things that we thought about as well. And this is a, you know, a genuine conversation that we should have is This makes sense because it creates that differentiator, that user experience. Like, it because if it's on desktop if it's on desktop, then this will be the ideal, interaction.
George Wang: Right? All these pieces are connected. Everything's feeding into each other, but but only if it's on mobile. How do you recreate that? Right? And there's not two different conversations. There's one conversation. So when you start Thing is full. You don't you don't you don't see the you don't see, you know, the clarity card. You don't see the scenario. So as you if you click this, and then it comes down, and as you're you know, like, it's like it's the the thing that I forgot, lost my train of thought around is, when when you have a conversation, that creates a scenario.
George Wang: So you're all you're on the scenarios views now. Scenario view? Then what? You just play around with scenarios. How are you getting back to the conversation? And we want this to be this continuous loop that will the AI will get more insights through the conversation, and and it it's like it creates that continuity, amongst these different parts. David?
David Kawata: No. I I love the awareness that we're trying to provide value, but not be disconnected on the conversation. So I understand the problem statement and the challenge on the mobile. So I I, again, I wanna commend you for coming with a plan to try to attack that on a small screen. And so I think, it'd be great to have the team have access to this to help the conversation to your point. How can we create the the transparency and the connection?
David Kawata: Value of information. So, it's a challenge, and I see the the process. And, again, the continuation of understanding the communication of the logic process of that you have so that everyone understands the starting point and the iterations you went to to get here. I think that would be helpful for the the team to provide value in a conversation. But I like the approach. I like the understanding of the challenge that we're trying to accomplish. So thank you for that.
George Wang: Thank you. I know we're way past time, so any last thoughts, before we wrap? So a couple of follow ups. One is, you know, work with the team to figure out, you know, how do we attack this. And so, you know, me and Charlie, let's have a conversation today. You know, there's the the the colors. There's the, the f. The orb. Right? The the the, you know, the the the the visuals the visualization of how the the orb is gets created and how that actually, Roam, maybe that's not what you're asking for exactly.
David Kawata: The part
George Wang: You're you're seeing more for yeah.
David Kawata: what what with the animation, the transitions, the thought process of how the cards and the magic moments are are moving, again, it's hard to go from a Figma to understand what that is. I like to understand and feel that, communicate that, articulate that. I I need more to that that conversation, please.
George Wang: Yeah.
David Kawata: Second, one one of the big deals that I that I'm missing is the personalization. Right? The Wayne family could benefit from this. Right? In context, individualization, that was a real big lead of the understanding.
George Wang: Mhmm.
David Kawata: And, there's been a a consistent, opportunity to improve that I wanna make sure I highlight as I see the UIUS conversation that there's an understanding. Oh, the Evergreen property can pull out 52,000. Like, how do we sprinkle the personalization so the user understands this is not a Roam? This is the conversation that's meant. Me. So I wanna make sure that that is taken away from the team,
George Wang: Yes.
David Kawata: please.
George Wang: Yes.
David Kawata: Thanks.
George Wang: Yeah. Yeah. I mean, one thing that we're learning from building agents is there's there's so many, like, pars there's You the tone, and then there's,
David Kawata: Yeah.
George Wang: like, the logic and, like, now it's how do we marry it all together? It's it's really it's, it's challenging, but it's really fun.
David Kawata: It is cool. So, like, when I go to Claude at 05:00, the golden hour. So Working at 02:00 in the morning. Hey. You're doing something late. Burning the midnight On-It. Those nuances upfront to say, hey, David, what we're working on, that is, again, AI based personalizations, timing zones, where we're at in a place. That is what I'm expecting to trace that connection to the thing and then get that out of public I want. So what this, this tool is gonna be available where?
George Wang: We we can put it up on, dev.apricot or, prettyhigh.apricot.dev.
David Kawata: Perfect.
George Wang: Yeah.
David Kawata: Thank you.
George Wang: Yeah. It's there's still, you know, many broken pieces with this and what, you know, what I would yeah. You know? So be careful. Going through it, but the, yeah, if you have any questions, let just let me know while we can put this up. Alright.
Vinicius Ferreira: Okay.
David Kawata: Probably, any last word?
Vinicius Ferreira: Okay.
Charlie Gann: Thanks, George. Well, I'll, I'll definitely throw some stuffover to you some some thoughts of thanks for sharing.
George Wang: Okay. Alright. See you guys. See you guys.
David Kawata: Thank you. Appreciate it.
George Wang: Bye.
David Kawata: Bye.