Watch modeSubscribed

TBPN · 5 min · Oct 5, 2026

Valon built a servicer first to prove its mortgage software

Regulated mortgage firms wouldn’t buy an unproven system, so Valon built a servicer around its software, then sold it after proving the model. Linda Du says AI makes modern systems of record and APIs essential.

LD

Linda Du

President & COO, Valon Technologies

  • John Coogan, Host · Co-host, TBPN

    Co-founded Soylent and Lucy; OpenAI bought his show TBPN in April 2026.

  • Jordi Hays, Host · Co-host, TBPN

    Founded Branded Native; co-founded fintech Capital, which Rho bought.

About this segment

Regulated mortgage firms wouldn’t buy an unproven system, so Valon built a servicer around its software, then sold it after proving the model. Linda Du says AI makes modern systems of record and APIs essential.

From “Minecraft AI Mods, Should You Sell Your Data, Construction Paper Really Fell Off”, Oct 5, 2026. Watch the full episode

“Mortgage companies are heavily heavily regulated and so they can't buy a de novo unproven system, it's just not something that from a risk tolerance standpoint they can do.”
LDLinda Duat 1:31:39 · TBPN1:31:39

Takeaways

  1. Mortgage firms were reluctant to adopt unproven technology.
  2. Valon operated a servicer to prove its software before selling it.
  3. AI agents need reliable records and APIs to work with legacy systems.

In this episode

Full episode
  1. 1:02:47Ridge rejected a small data deal because the risk outweighed the return4:59
  2. 1:41:49Neo is concentrating its bets on fewer companies and people4:04
  3. 1:50:56AI labs could face liability when their models choose illegal acts3:43
  4. 1:31:13Valon built a servicer first to prove its mortgage software5:18

  5. 1:07:23AI shopping agents may take fees from both sides4:11Up next
  6. 1:15:00Ridge generates quarterly reports with three AI prompts2:44
  7. 2:07:03Today's frontier models will run on devices in six months2:29
  8. 2:17:20RL environments roughly double in complexity every 6–8 months2:58

Transcript

From the transcript TBPN publishes with the episode.

  1. 1:31:25

    Walk me through the the verticalization of your business and then what it seems like is de verticalization because you sold your servicing operation but partnered with Carrington Mortgage Services to use your your software. Right? So what was the what was the back and forth there strategically?

  2. 1:31:44

    Yeah. So the plan from the start has been that we wanted to deploy software and now AI into the industry. Mortgage companies are heavily heavily regulated and so they can't buy a de novo unproven system, it's just not something that from a risk tolerance standpoint they can do. And so, the way that we decided to go to market, it took six years, is we actually built a servicer first. Sure. Built the technology alongside that servicer and then when the market made it clear that we've kind of proven our point, We sold the servicer and today, we're just a pure play.

  3. 1:32:22

    Yeah. What does the actual customer experience look like? Does this get white labeled into like bank software and then it's rolled up in in, like, your bank dashboard? Or or are you handling, like, mailings and notifications, dealing with defaults? Like, how what... How how broad is your mandate in terms of the relationship that a a lender has with a borrower?

  4. 1:32:49

    Yeah. So, I I think of servicing as basically all of the work Mhmm. Behind any long running financial transaction. Okay. So, mortgages, perfect example, it's a thirty year obligation. Yeah. And because it's thirty years, a lot of stuff has to happen after you've signed that initial mortgage contract. Payments have to be collected, escrow, tax, insurance has to be handled, a lot of things can happen to the homeowner over thirty years. They could, you know, fall behind on their mortgage, they could get married, someone could try to assume the mortgage, they could lose their job and need some help staying in their home. So all of that end to end for the life of the financial application is is servicing.

  5. 1:33:31

    So now we're fully... We're firmly in the AI era. Is... Do you even need to communicate the benefits of AI to your customers? Are there particular features that have unlocked or is it all just like table stakes and you are solid as like a system of record as a software product and the value is clear and so everything else is just sort of the cherry on top?

  6. 1:33:55

    Yeah. It's it's kind of interesting because the the industry today really runs on a system of record that was built in the nineteen sixties. So actually before the internet was invented.... There we go.

  7. 1:34:09

    Some people like they like they like cars from the sixties. Some people like... Technology. Vintage technology. Might be making... Yeah. No. The the artisanal vintage... Yeah.

  8. 1:34:21

    Yeah. Exactly. And for decades that was fine actually because humans could just use it, there were issues, but everyone was on a level playing field. Mhmm. AI has actually helped us a lot because now everyone feels like it's essential turns out that if you don't have a system of record that is a single source of truth with very basic things like APIs and tools that an agent can call, it's Yeah. It's kind of dead in the water.

  9. 1:34:46

    Yeah. Is there is is there movement or does there need to be movement at like the government level? The only thing that I've seen recently on the mortgage side is like the proposal for fifty year mortgages, which I assume would be bullish for you actually. Right? But but is there is there any like rumblings? Obviously, in Washington, everything is about like AI safety. But is there is there any movement for updating like any federal level regulations or rules to just make the actual experience of getting a mortgage as an American easier in the modern era or is this something that can be purely handled by the private markets?

  10. 1:35:27

    Yep. Purely handled by the private markets. Okay. Obviously, I mean, the thing about servicing is that if it's very simple, you don't actually need complicated technology Sure. To do it. It's, you know, in some ways it's... You can think about it on two dimensions, right? Mhmm. It's like I can go out and buy a coffee on the and that's actually a very simple experience. I don't need to, you know, back in the olden olden days, like, I don't know, bring some chickens and bar the machine. Now you need an

  11. 1:35:53

    AI agent to call a thousand coffee shops I'm going try and reserve you a slot I'm gonna try to pay for a latte with a chicken to And you need to pay for Just to see if we still know how. Yes. Barter economy. Yeah. Just just...

  12. 1:36:06

    Yeah. Exactly. Let me know how that goes. But but if you have a longer time horizon and more complexity complexity in in the the form of regulations, then that's where you actually need sort of purpose built technology. Yeah.

  13. 1:36:18

    Yeah. That makes a ton of sense. Well, congratulations on the round.