Breaking into the twenty-year startup graveyard of mortgage lending | Naren Krishna

Balerion AI is an agentic AI platform for mortgage loan manufacturing, automating origination work like document analysis, income review, and underwriting for lenders. The company came out of stealth this spring and is backed by a $6M seed round led by Kleiner Perkins, with its platform in production at FM Home Loans, a lender originating two billion dollars in annual loan volume.

In a recent episode of BUILDERS, we sat down with Naren Krishna, CEO and Co-Founder of Balerion AI, to learn how the company is breaking into an industry he describes as a twenty-year startup graveyard.

Topics Discussed: Why lending has been a startup graveyard for twenty years, and the three failure buckets Naren sees in past mortgage tech companies Why the cost to originate a loan went from about $4,500 in 2007 to around $13,000 in 2026 even as technology improved Why Naren argues the worst thing that happened to the mortgage industry was its vendors, with lenders using an average of 32 point solutions How 100 pages of guidelines became an 800-page PDF that changes every couple of months, and why people spend, not technology spend, absorbed the complexity The stare and compare problem: multiple roles reviewing the same 800 pages How Balerion recruited by marrying AI talent, infrastructure talent, and lending expertise What made FM Home Loans the right pilot partner, and how the case study opens the rest of the market The launch video that led to hiring a VP of sales off a LinkedIn like His color-coded framework for executive updates How the vision expanded from automated underwriting toward loan quality for capital and secondary markets

GTM & Technology Adoption Lessons:

Study the graveyard before entering it: Naren puts past mortgage tech failures into buckets: companies beholden to macroeconomic shocks because they served only one loan category, companies that forced their own UI or database onto lenders instead of living where underwriters and processors already work, and companies that never understood the distinct needs of independent mortgage banks, depository institutions, and the secondary market. Sell a generalizable system, not a point solution: Lenders use an average of 32 vendors for one manufacturing process, which means underwriters must learn the process, what each vendor does, and how to fill the gaps between them. Pick a pilot partner who already believes: FM Home Loans wanted a world where a mortgage application works like a credit card application, knew it lacked the in-house AI expertise to build it, and originates two billion dollars in annual volume across a broad mix of loan types. Buy credibility in relationship-based industries: Balerion's VP of sales has been in mortgage for decades and opens doors the technology alone cannot. In an industry where everyone claims AI, relationships get the meeting and the technology has to back up the claims. Marketing bets pay off in unpredictable ways: The launch video cost about ten grand and could not be tied to a tangible outcome, until someone liked the LinkedIn post about it and Balerion recruited them. Simplify communication to match attention spans: Naren color-codes executive updates green, amber, and red. Green means do not even look at it; red means you will get a call from me in four hours. Hire against the bet, not the present: Naren spends a quarter to a third of his time on hiring and posts JDs six months ahead of anticipated need, because finding the right person takes three to four months.

// Sponsors: Front Lines -- Silicon Valley's leading Podcast Production Studio. We help B2B tech companies launch, manage, and grow podcasts that drive demand, awareness, and thought leadership. Mention you are a listener and get a 10% discount. www.FrontLines.io/Podcast-as-a-Service