Speaker: You’re listening to a new episode of The Brave Technologist. Our team returned from December from New York where we attended the SEC AI Summit for the second year in a row where we created a popup studio on the show floor and had the privilege of sitting with some of the conference’s top thought leaders in AI who were speaking at the event today, you’ll meet Leah Price who serves as Vice President leading the 10 man AI platform at better the AI powered home ownership company.

Under her leadership, the 10 Man AI platform will be offered to lenders and brokers across the country as an API accessible platform. In this episode, we discussed how Tinman embeds generative AI across the entire Mortgage Tech stack, the limitations of traditional ROI models in enterprise AI adoption and their vision for outcome-based pricing, how AI avatars and bots reduce borrower friction and improve trust, and what the future of mortgage lending could look like with AI as his infrastructure.

And now for this week’s episode of the Brave Technologist.

Luke: [00:01:00] Leah, welcome to the Brave Technologist. How are you doing today?

Leah: I’m really good. How are you?

Luke: Great, great. you spoke yesterday here at the AI Summit and kind of looking back on your session, what was a key takeaway you wanted to have people to walk away with from your talk?

Leah: Yeah. in preparing for my session, it was on ROI mm-hmm. In ai. I. That what we have with Tinman AI platform mm-hmm. And our pricing mm-hmm. Really is a game changer.

So we price based on outcomes. Mm-hmm. So most technology in general, and AI right now is priced on a subscription basis for a per seat basis, but we’ve completely flipped that on its head.

We’ve got our software solution that has embedded AI at its core. Mm-hmm. And we only charge our clients. When they realize a revenue event.

Luke: Oh wow. That’s awesome. Yeah. Yeah. And

Leah: so I was asking the audience, you know, of you all, are you buyers [00:02:00] of ai? Are you builders or are you sellers? and they were all distributed, but I said, you know what?

it’s in your power to tell your vendors, Hey, what if I don’t pay unless I see an outcome?

Luke: Yeah.

Leah: And that would really flip things upside down.

Luke: That’s really awesome. Like, you know, what do you think is one of the, the biggest misunderstandings around how people are measuring. Success. ‘cause obviously I think, with what you’re doing, there’s a clear ROI outcome, but are there just any general misunderstandings you’re seeing?

Leah: Well, what I realized yesterday was that this whole concept, the traditional way of thinking about ROI at a big enterprise, and I, come from the federal government uhhuh, where in order to kick off any kind of technology implementation, you’ve gotta thoroughly vet what the cost is gonna be and upfront predict what the, either the reduction in costs or the.

The new benefit would be mm-hmm. And now I’m thinking that whole paradigm needs to completely shift And what if we just didn’t think about upfront investment at all? what if [00:03:00] everybody in software just said, no, I’m just not paying until I see some kind of benefit?

Luke: Yeah. I mean, I think it’s not totally foreign either, right? Like a lot of people have free legs of, software. People are kind of used to that. We haven’t really, we’ve seen it kind of limited on. AI side, but it seems really interesting in your guys’ application. Almost like a cost per action or something like that.

Leah: Exactly. Yeah. I mean, on the AI side, what we’ve thought about and what we see, so in general, just software, when you add on some kind of an AI chatbot or a bell and whistle, a vendor will try to charge up for that

Luke: uhhuh

Leah: and say, oh, here’s my agent and let me upsell you on this agent. We’ve decided we’re not doing that.

Luke: Oh, that’s cool.

Leah: It’s the same price. Whether or not you use. Use our chat bot on top if you use our voice bot. We also have human based avatars mm-hmm. That sit on top of our platform. All of that is embedded into the same price, which you only pay when you realize value.

Luke: That’s cool. Why don’t we get into that a little bit more?

Um,you’re leading this effort around the 10 Man AI platform. You [00:04:00] know, at better, what is Tinman doing and how is this changing the Mortgage Tech landscape?

Leah: Yeah, okay. And I won’t assume that you know anything about the Mortgage Tech landscape. It is super. Old school. Yeah. And I’m gonna talk about it both from my perspective now at Better

Luke: Uhhuh,

Leah: but also from my perspective, having regulated the industry.

Yeah. Prior. So the mortgage industry 85% of lenders are on a software stack that was built in the 1990s.

Luke: I believe it. it’s one of those things where, I mean, I think I remember doing like, refinancing, right? Like in 2019 or 20 in this late. Came to our house, the notary with a stack of papers, that was like half a ream tie that we had to sign through.

So it’s kind of one of those things where it seems ripe for disruption. Right.

Leah: Totally. Yeah. And so when I was at working at the federal regulator, I was leading an office of ai and so I had a lot of interaction with people in the industry. Mm-hmm. And lenders. So both vendors and lenders. And the one lender that was [00:05:00] out there at the bleeding edge of leveraging technology and generative AI was better.com.

Mm-hmm. And so I had them on my radar. I was like, wow, these guys are really like, they’re gonna kill it. Once gen AI tools, like really take center stage, which in the mortgage industry would take a while, but they were so ahead in embracing it, I was like, you know what, that would be the place I would wanna work.

If I would leave a federal government. But there. A lender, and I wasn’t really into being a lender, but at some point over the course of the last year, the CEO Vishal Gar, who is like a Steve Jobs kind of Oh, cool. Genius. He decided he wanted to license out the software to other lenders. Okay. So we are kind of in this AWS moment, Uhhuh for the mortgage industry.

Luke: Interesting. Yeah.

Leah: So you think of Amazon’s selling book. And then it shifts to becoming the infrastructure for all of these major enterprises.

We are making that shift from being a lender serving [00:06:00] borrowers, to now being the infrastructure for all of loans everywhere.

Luke: That’s awesome.

Leah: And so the mortgage industry itself, if you think like in terms of value, it’s like $2 trillion Yeah.

Worth of mortgages are originated per year.

Luke: Yeah. That’s a huge.

Leah: Absolutely. And most of those lenders are on that, infrastructure that was built in the 1990s, right? Like this is the opportunity of a lifetime for me. And so that’s, where I’m sitting right now. I’ve got this spectacular piece of.

Software that is just making the pivot. We’ve got three clients that we are implementing right now. Mm-hmm. And so, that’s about to get a lot of news in the world as we go into broad launch with the software.

Luke: So is that, I mean, I can imagine too, like you’ve probably got a new generation of home buyers, right?

Like that, and in the home is still that kind of big asset, you know, American dream type of asset, right? Like for people to get into kind of socially move forward in a lot of cases. is this process just [00:07:00] making it easier to buy or do you get a lender, like how does it work?

Leah: Yeah. Okay. So, baseline typically there are like eight different systems Yeah. That a lender uses to manufacture a mortgage

Luke: uhhuh.

Leah: So there’s a point of sale that the borrower interacts with. Then there’s a loan origination system that internal users, that a mortgage lender work with. Then there’s pricing engines.

There are all kinds of different pieces. better.com had built its own end-to-end platform. Oh, okay. So one single software stack. So what that teed up very nicely for us was when Gen AI became all the rage.

Luke: Mm-hmm.

Leah: Competitor vendor products, they could only add Gen AI into their one little piece. Mm-hmm.

Of the puzzle. Puzzle. So the borrower could talk to a chat bot,

Luke: Uhhuh,

Leah: and then the loan origination system had some kind of bot for internal people, et cetera, et cetera. We’ve been able to do is layer gen AI on top of this end-to-end system. Mm-hmm. So nobody else has this [00:08:00] one bot, which is available in voice or chat or avatar.

Nobody else. Ours is called Betsy, by the way. Better Betsy. She’s able to access a loan file, interact with the borrower, Uhhuh, do phone calls. All of that information that she can collect about a borrower is then also in the context of what a. What an internal user at a mortgage lender has access to.

Luke: Oh, cool.

Leah: And then Betsy also has access to everything on the pricing side. SoHow much does a loan cost? How much is a lender gonna make on that transaction? And so it’s really special what we are able to offer now in terms of an AI offering.

Luke: Yeah, that’s great. I mean, there’s so many pieces, like you’re saying, and, and you keep hearing people talk about genic things and it seems like such a good area to kind of show that off.

I mean, where do you see the biggest opportunities for AI to reduce friction? and not just automating tasks, but in, in general, in home ownership and lending.

Leah: Well, we’ve really leaned hard into [00:09:00] automating tasks internally at better Uhhuh. So that’s like our, like we’ve seen all kinds of gains.

Our loan officers are able to handle three times the volume Oh wow. Of regular industry underwriters and loan processors can handle 10 x the volume. So I don’t want to like. Poo poo or say that’s not cool ‘cause it’s really important Right. For what I’m selling. But in terms of gang changers, in terms of real friction for borrowers you think about like we’ve been experimenting with an AI avatar.

Mm-hmm. So I have Aaliyah, there’s Aaliyah avatar out there who you can interact with 24 hours a day.

Luke: Uhhuh,

Leah: she has a whole knowledge base. Of information. She’s not trained with my personality,

Luke: so we don’t wanna interact

Leah: with that.

Luke: Oh, come on.

Leah: But we have another one named Ryan, who’s trained who’s been fed all of his personal quirks and his sayings nice and his friendliness.

And he can provide really custom ter tailored relationship building with a borrower. So all of that [00:10:00] being said, the, the game changer there is 24 7 a borrower can. Really have access to mere human grade interaction really, with these avatars. Wow, that’s awesome. So yeah. So you think about sometimes borrowers are embarrassed about their questions.

Luke: Yeah. Oh, totally. Yeah, yeah, yeah, yeah.

Leah: And so maybe they don’t even wanna really talk to a human, like they want the essence of a nice person. Right. But they don’t wanna actually be, feel like they might be judged.

Luke: Yeah. Yeah.

Leah: And so there’s a whole open. Opening there, I think for borrowers to feel more comfortable.

So that’s one of the things that I’m really excited about. That’s something

Luke: I hear echoed all over the, I’ve talked to all sorts of different people, people in the academic space tell me the same thing where it’s like, look especially for people who maybe new to the country, like they feel uncomfortable asking questions.

it just seems like a great way you can ask every like, quote unquote dumb question to the, the bot and get the answer back and kind of go in. Or people, like, some people might not even like go forward with the process. Because they’re unsure about something. Right. Exactly. Like, Yeah. That seems really interesting.

Leah: Yeah. [00:11:00] And so like my avatar is really funny. I cannot be that funny. I’m not capable. But you think like she will listen to every single word that a borrower says. A human loan officer can’t focus that well, a human loan officer can’t possibly process pricing. Right in real time. But Betsy can, because she has access to that pricing engine.

And actually we’ve programmed, here’s another thing that is special about the way we’ve done this is that our gen AI tools, they’re not able to hallucinate, they don’t do any calculations. Okay. Because we already have this platform underlying that performs all the calculations itself. Oh, nice. So the gen AI is only calling the service.

Yeah. And that’s how it will provide any kind of estimate or calculation on the value of the loan. Or crisis

Luke: I would imagine too. I mean, just kind of with your background, looking at it from kind of working for the people too. you get a better sense of a, like, like a log of events, right?

Like, ‘cause I mean, you know who, you know, when the financial crisis happened, there were all these [00:12:00] stories about back rooms and Opaque processes and stuff. But it seems like here at least you have like kind of an audit trail or something. Like, is that, fair? Exactly.

Leah: So you cannot hide It would be very hard for, so typically lenders record conversations.

Mm-hmm. We have full transcripts of all of our bot interactions. Yeah. And then we have a guardian model that’s constantly monitoring those conversations. Wow. And so if a conversation is going south for whatever reason, if a borrower is having an issue. You the guardian model will cut it off and intervene and get a human.

Wow. So you can’t do that with human beings. Yeah. And unfortunately, you can’t actually really trust human beings to say the right thing. We were reviewing a transcript recently of a borrower whose husband had died. Oh. And they were talking about that difficulty from the financial transaction perspective.

But the borrower in the midst of this really hard conversation then said, oh. Know what, I’m sorry. I don’t have that information right now ‘cause I have chemo brain. Oh. And [00:13:00] then you’re like, oh my god, And the bot picked up on that immediately.

Luke: Really?

Leah: And was so empathetic, you know, like a, a human, it can be hard for a human to know what to say.

Yeah. And can you expect that humans are gonna be more sympathetic?

Luke: Right. Right. But

Leah: the, but Betsy, the bot, she said the perfect thing and then transitioned and said, you know, if you need more time, just let me know. So like we can code these bots to say the exact thing that we want them to and actually put more thought into how we want to treat our customers.

You can also customize one of the things that we’re letting our customers do is rebrand the whole software mm-hmm. To whatever they want. Oh, cool. So, let’s say Brave wants to become a loan originator, right? We could call the. Whole system. Brave. You could, I don’t know. It’s your, it’s a lion, right?

Right. Yeah, yeah. You could say, Lionel is your chat bot and we could tailor his voice and we could give him SAS an attitude. That’s awesome. And we could [00:14:00] tell him, I want you to handle borrowers, you know, with a little bit of attitude.

Luke: Yeah, I can imagine. I mean like, ‘cause you have home buyers of all different demographics and makeups and all that stuff, and the ability to kind of tailor it.

It’s just like, because those are the things where people kind of get thrown, by bad behavior. Right? where people kind of. Being deceptive or whatever, but if you have a system that can kind of be tailored to go towards a certain type of buyer in a, in a way that’s productive, it seems really interesting.

Exactly. what kind of shifts in borrower expectations are you seeing as AI gets woven into the process?

Leah: Yeah. So, borrowers are really starting to demand that they be met where they are.

Luke: Okay. And

Leah: that’s good for us because we’re leading the charge and having these gen AI tools who can, call mm-hmm.

Luke: Can meet.

Leah: On Zoom or who can text. Mm-hmm. So over time, I expect borrowers to get more and more demanding, which will be great for us. Awesome. Also, you know, I talked about custom tailoring our bot to meet whatever needs. I think consumers are gonna get more demanding about having custom tailored [00:15:00] interactions.

Luke: Yeah.

Leah: So, you know, with a human being, let’s say, I don’t wanna talk to a dude like I, you know, I can’t demand that

Luke: as a consumer,

Leah: but as. If I’m interacting with a bot, it’s super easy to be like, you know, I would feel more comfortable if it was a woman in her fifties who is kind of maternal. Yeah.

Yeah. And that’s the persona. And then there you go.

Luke: It’s really interesting to think about too. ‘cause we, even, like in our products, we let you kind of tune the responses from our, our interface to how it is. You can totally see that being something you could maybe pour over or something like that.

Exactly. And kind of have that continuity in, cause it’s really interesting like talking about this ‘cause it’s like I’m hearing a lot of care and depth around like treating a human, like the human right, like having the right human responses with this stuff. I think that’s something people kind of worry about with these things.

It’s just like that it will be too cold or, you know, mechanical. So it’s really

Leah: cool. Ooh, I like my ai cold mechanical.

Luke: I don’t want any

Leah: fluff, man.

Luke: Hey, everybody’s different. This. Um,

Leah: So, and one thing that we’re looking forward to, and I think borrowers are gonna start to demand, like [00:16:00] everyone’s really comfortable with chat, GPT and that kind of interface.

Mm-hmm. Soon it’s gonna be, I’m a borrower, I want a house. I don’t want to think about the loan.

Luke: Yeah.

Leah: I’m just gonna go into chat GBT and say, I want to buy a place on the, in the West Village. Yeah. Make it happen. Yeah. And then chat GBT is like, oh great, I got your stuff in Dropbox. I’m gonna go ahead and make your loan and borrow.

Doesn’t even have to think about who the lender is. Yeah. We wanna be the infrastructure for, that’d great. Great. So we’re working with OpenAI.

Luke: Oh, awesome. On that concept. That’s great. I mean, seeing that work really well with like travel and even with auto buying, right? Yes. And it’s one of those things where people.

People really hate having to haggle or go through, you know, lots of bureaucracy on that stuff. So it’s, it’s really interest. Yeah. how do you see that mortgage experience? Is that kind of how you see it going like five years from now or I mean, if you could wave a wand and be five years in the future.

Leah: Yeah. I, think it’s a big open question. What is the front door for borrowers uhhuh? But I know that zero borrowers sit around thinking, oh, I’m excited for my to [00:17:00] talk to a loan officer. Right. Or an originator. And I. Think most of them don’t wanna think about who the lender is. Mm-hmm. Mm-hmm. They just want the house.

Luke: Right. Right.

Leah: And so how can we best allow a borrower to focus on the object of desire

Luke: mm-hmm.

Leah: And not get caught up in all the other junk.

Luke: Yeah.

Leah: And so that front door I think is gonna be however borrowers are already more comfortable ordering other stuff.

Luke: Sure.

Leah: So like today, people are used to ordering things on Amazon Uhhuh.

it feels like in the future everybody’s gonna have some kind of cha pt like interface Yeah. That knows everything about them. Their preferences is customized to them. I suspect that that is the future.

Luke: Yeah.

Leah: And our software will just be integrated there and it’ll be completely seamless. And then, you know, then it’s not just mortgages, then it’s all types of financial products.

Mm-hmm. Mm-hmm. Auto personal loans and then really global penetration. Is much easier. Yeah. Right. Yeah. Imagine that [00:18:00] distribution channel is just ChatGPT, or whoever is more

Luke: Yeah. I mean, and there’s been a lot of like innovation too around, like things like Zillow and, and stuff where people are out there looking online for these things and it’s almost like, okay, you know, seeing the back end of it now, being able to kind of get an easier process to getting the keys to your house or whatever, you know, like it’s really makes a lot of sense.

So what role do you see standards bodies like SMO playing as AI becomes. More embedded into, into underwriting and decision. Maybe we can say what Ms. Mo is. Nobody ever

Leah: asks me about Ms. Mo, but I’m a huge fan of Ms. Mo. I’m a senior advisor Tommo. I’ve led the emerging technology community of practice.

So I’ve talked to them a lot. About what their role should be in the industry, and it’s really around data standards. Mm-hmm. If we can have a unified view of even what defines a given metric mm-hmm. Or a given value that’s huge value so that everybody’s working off the same data.

Luke: And for the uninitiated, so Ms.

Mos, is they like kind of like a standards body around these or, yeah. Okay. So

Leah: the Mortgage Bankers [00:19:00] Association is the trade association for the mortgage industry. A subsidiary of that is the Mortgage Industry Standards Maintenance organization. Got it. You got it. Yeah. So, so we have been a gold sponsor and in prior roles I’ve also been a gold sponsor.

Oh, great. They’re very important.

Luke: That’s awesome. Shout out to them. so tin Van’s being offered as like an API accessible platform. What do you think the great AI infrastructure for lenders and brokers will be in the next few years? This is kind of like what you’re talking about around Yeah.

Just being able to integrate.

Leah: Yeah. I mean, as I said, 85% of the industry is on a platform that was built in the nineties. Yeah. Yeah. So the future is theirs. All they have to do. Is make that change. And the way we’ve set it up, there’s no implementation fee, no upfront cost, no per seat. If you get implemented and you don’t use it, we don’t make any money.

Wow. So we are completely incentivized that our customers make money.

Luke: Mm-hmm.

Leah: Which is that that really puts our feet to the fire. Yeah. That we have to deliver value.

Luke: Yeah.

Leah: So I believe that [00:20:00] that paradigm, I think in software that is disruptive. The mortgage industry, it’s completely revolutionary. Yeah,

Luke: I think so too.

Especially based on what we’ve been talking about here today. It’s like, you know, it doesn’t sound like there’s a lot of competition either with the space or people doing it in the same way that you all, I mean you’ve got, you know, guardian agents, all these other things like already in place.

It’s really, really interesting. You might have to do a couple rapid fire questions. Sure. Like at the end that we wrap this up. So, so what’s a, the mortgage step you’d eliminate forever if you could?

Leah: New York City co-ops.

Luke: Alright. What’s the most underrated metric for judging AI Success?

Leah: Delight.

Luke: Okay, excellent. And what’s one word to describe the future of home ownership?

Leah: Piece of cake.

Luke: Excellent, excellent. thank you to, you’ve been really gracious with your time. I think our audience is gonna learn a lot from this. where can people find out more about what you’re up to or, or reach out to you or, more about what you’re doing at Better?

Leah: Yeah, so better.com is the lender’s website better.com/tinman to. Find out a little bit more about Tin Man and to play with our avatar.

Luke: Awesome.

Leah: And [00:21:00] then I’m on LinkedIn, Leah Price.

Luke: Excellent. Well thank you Leah. I really appreciate you making the time. It’s been a great conversation. I’ve learned a lot and I think our audience has too.

Leah: Thank you so much.

Luke: Excellent.

Thanks for listening to the Brave Technologist Podcast. To never miss an episode, make sure you hit follow in your podcast app.

If you haven’t already made the switch to the Brave Browser, you can download it for free today@brave.com and start using Brave Search, which enables you to search the web privately. Brave also shields you from the ads trackers and other creepy stuff following you across the web.