Better founder Vishal Garg details Tinman AI platform shift

Article Summary

Better founder Vishal Garg said Tinman AI is driving about 60% of company activity as Better pivots from ad-driven DTC lending to partners. He also outlined a $60 million cost reduction plan and a focus on contribution margin profitability.  AI Summary

Better founder Vishal Garg sat down with HousingWire President Diego Sanchez on the PowerHouse podcast today to unpack his proxy victory, the company’s shift away from ad-driven direct-to-consumer lending and a plan to turn its Tinman AI platform into shared infrastructure for brokers, IMBs and fintech partners.

In the wide-ranging conversation, Garg talked first about his proxy fight to return to the company he founded.

“The odds of a founder CEO coming back and removing the entire board — that happens 1.7% of the time successfully in the history of proxy fights. So, I didn’t know that going in, apparently, but like 60-to-one odds against,” Garg said.

Garg was removed as Better’s CEO in early August, replaced on an interim basis by Daniel Lewis. Garg then launched a campaign to replace members of the board, winning over enough shareholders to regain control of the company.

He says shareholders backed him because they buy into Better’s mission to make homeownership “cheaper, faster and easier,” and because they see early results from Tinman AI, the company’s mortgage manufacturing engine.

In his short time away, Garg said he reflected on the mistakes he made as CEO. “The biggest thing I did wrong was not push the AI hard enough. You know, really just still kind of kept trying to convince people that this path was the best path, trying to convince people that the agent needed to be perfect, right? It did not need to be perfect.

“All of those things, big mistake. While I was doing that, there are other people out there, you know, 23-year-olds disrupting entire industries with a completely brand new perspective and just like letting the agents run.

“And so while we have this massive competitive moat of licenses, and over a hundred billion dollars of transactions on our platform that serves as learning data, the ability to rep and warrant — and all these sorts of things that your traditional 23-year-old coming out of Silicon Valley with a a mortgage startup doesn’t have the ability to do — we would very clearly lose that edge over time if we just didn’t run fast enough.”

Tinman AI and business mix shift

Garg said Better has raised “almost $2 billion” to remake the mortgage manufacturing process and acknowledged the company has “lost more money than anyone else” in the space. He compared that to the long burns at Uberand Airbnb, positioning Better as early in an effort to reshape what he called a $4.5 trillion industry.

Tinman — originally built to codify agency and investor rules from 2016 to 2022 — is the foundation of that strategy. Garg said the system delivers a “six basis point critical defect rate,” which he called the lowest in the industry, and that overlaying large language models has accelerated the process and stripped out human friction.

Better’s revenue peaked at about $1.5 billion during the refi boom, then dropped to roughly $75 million in 2023. Garg said the company is now at a “$220-plus” revenue run rate, with about 60% of activity tied to Tinman AI.

Crucially, he said Better no longer cares “about the customer interaction surface.” Instead, the goal is to plug Tinman AI into whatever front end the consumer, lender or broker already uses — whether that’s a fintech app like Credit Karma, an AI agent interface, a mortgage broker’s POS or a branch loan officer’s tablet at an open house.

“We are not even seeking to disrupt the industry anymore,” Garg said. “We’re seeking to empower the entire industry and uplift the entire industry by taking Tinman and making it available to anyone and everyone in the industry to power their business.”

From DTC to partner platform

Garg framed Better’s first phase, from 2016 to 2023, as a DTC-heavy play. The new phase, he said, is “interaction surface agnostic” — with growth coming through distributed retail, broker and fintech partners instead of Google and Facebookads.

He pointed to Better’s relationship with NEO Home Loansas an example of how Tinman can sit behind distributed retail brands. Garg said Neo’s loan officers went from averaging roughly 2–2.5 loans per month to closer to 5 loans per month on Tinman, with higher commissions because they can handle more business and deliver more consistent execution.

He said Tinman reviews decisions across “all 65 investors” before a loan is rejected, reducing fallout and helping attach additional products like HELOCs. The strategy, he argued, is similar to Uber’s path, where local operators plug into a central platform rather than run their own full tech stacks.

“The Uberization of retail is what happens,” Garg said, describing a model where Better supports multiple distributed retail brands and expands city by city.

Wholesale and home equity

Garg also confirmed that Better is already operating as a wholesale lender.

“We are a wholesale lender,” he said, noting the company currently has about 40 mortgage brokers on its platform, focused today on home equity.

He described that as a “land and expand” strategy. Over time, he envisions brokers evolving into broader “home advisors” or “loan advisors” who can also sell insurance and maintain long-term relationships with borrowers.

Garg highlighted that past investments — including title, homeowners, auto and life insurance entities under the Better Cover umbrella — didn’t produce strong unit economics when driven by Better’s own human salesforce. But he said those capabilities can make more sense when product distribution runs through brokers and retail partners who already own the consumer relationship.

“It is so powerful to tip the platform in the hands of people that know what they’re doing,” he said.

On where Better is pulling back, Garg was direct: paid, ad-driven DTC.

He said Better has already “cut D2C marketing” and questioned the logic of spending “two and a half million a month for buying clicks on Google” when he was able to mobilize tens of thousands of proxy voters himself using low-cost AI video tools and social platforms.

“I am short ad-driven D2C, brand-driven D2C, because it will die a fast and rapid death in the age of the AI agents,” he said.

He contrasted that position with lenders that have invested heavily in mortgage servicing rights and large workforces, saying those business models are effectively “long consumer friction” at a time when AI agents such as Muse, Instinct and others could make it easier for borrowers to move and refinance.

“You cannot charge people 50 basis points on a mortgage and hope to make up for it because you can bid more than everyone else on Google,” Garg said.

For lenders that still lean on digital ad spend and lead buying, his stance underscores the risk that acquisition costs continue to rise while AI tools make rate and fee comparisons more transparent.

Cost cuts and the profitability push

Garg acknowledged recent layoffs at Better and called the current cost structure “unsustainable.”

He said the company needs to reduce costs by about $60 million and has identified or executed on roughly $24 million of savings so far. The goal, he said, is to move quarterly operating costs from the mid-$60 million range into the $40 millions.

“For us to be the platform for the industry, for us to be the AWS for the industry… the industry has to know that Better will always be there,” he said. “And the only way the industry knows that Better will always be there is if Better makes money every month.”

Asked whether Better will be profitable a year from now, Garg said the “core parts of the business will be contribution margin profitable,” unless the company chooses to lean into large growth or acquisition investments. He differentiated this plan from past profitability promises by emphasizing that this time Better is attacking both revenue and expense lines.

“In the past, we were keeping expenses the same and using AI to help power growth,” he said. “Now operating costs are going from that number down by 25 million… and we’re focused on still growing.”

Garg also argued that Better trades at about 1.5x revenue, a steep discount to private-market mortgage fintech valuations he said can reach 10x–50x revenue. Reaching profitability, he said, is the “best way to remove that discount” and create value for shareholders.

Leadership and CEO search

As part of his return, Garg has committed to hiring a CEO. He said he is looking for an executive who is “AI pilled,” has a deep understanding of credit risk and has experience building large B2B “outcome as a service” platforms — delivering specific outcomes instead of just software licenses.

He argued that mortgage is an industry that “blows itself up every 10, 20 years” and that the next CEO must be grounded in how credit cycles and product standards can go wrong.

Garg said strong people skills are also critical — and acknowledged that is an area where he has historically struggled and needs complementary leadership.

About Sarah Wheeler
Sarah Wheeler is Editor in Chief of HousingWire, where she sets news coverage strategy across mortgage finance, residential real estate and homebuilding. She hosts the award-winning HousingWire Daily podcast, the industry’s most listened-to housing news podcast at more than 1.4 million downloads a year, and writes on federal housing policy and the Federal Reserve.see full bio

Most Popular Articles

MLS executives weigh Open Real Estate and listing control

MLS leaders respond to Open Real Estate, a national listing sharing effort with a one business day rule, and debate control and value.

• Homebuilder profits, why brand beats volume for Toll Brothers 

  Oct 30, 2026By Scott Finfer

• A path to sub-7% mortgage rates is back on the table 

  Oct 31, 2026By Logan Mohtashami

• Exclusive: CrossCountry Mortgage demands payment from former LOs, invokes 2-year non-solicitation 

  Oct 30, 2026By Neil Pierson

• Strata Land launches as builders face land-capital squeeze 

  Nov 02, 2026By John McManus

• D.R. Horton’s silver-lining playbook for 2027, … with a catch 

  Nov 02, 2026By Dan Oppenheim

Latest Articles

Why yellow pad math still beats the spreadsheet in builder deals

[In the third of seven excerpts from ‘Less Wrong, More Right,’ Scott Cox argues that models can hide what matters by weighting every input equally. Simple back-of-the-envelope comparisons can identify what must be true for a homebuilder deal to work and where opportunity may be underpriced.] One of the most impressive people I got to […]

• A $20M fine stalks Los Angeles over delay of a new housing project 

• Real estate is moving toward more data privacy. Appraisers say UAD 3.6 is moving the other way 

• Could dual licensing grow as agents, LOs seek revenue opportunities?