What Happens When You Build Software for Agents

a16z Deep Dives · 2026-10-08 · 64 min
https://www.youtube.com/watch?v=TyWe_QQtYEUVideo summary
Vesta CEO Mike Yu says its mortgage platform can automate 75% of origination work, while PennyMac reported 25% lower operating costs.
On the a16z podcast, Kimberly Tan interviews Vesta co-founder and CEO Mike Yu about replacing mortgage lenders’ legacy loan-origination systems with a modern system of record and AI agents. Yu says outdated databases—with roughly 30,000 fields and single-writer limits—block automation, while Vesta’s structured platform can run agents in parallel. After founding Vesta in 2020, the team funded its first loan in December 2022; Yu estimates the product is now about 75% of the way toward automating possible back-office work, though lender trust and government access remain hurdles. PennyMac said it had cut operating costs by 25% before Vesta deployed its agents. Yu also argues long-term rates may stay high due to national debt, AI capital demand, and uncertainty, and says vertical SaaS can endure because regulated mortgage workflows demand specialized software and implementation support. He sees 30-year fixed mortgages as vital to the American middle class and says Vesta’s mission is to make them less costly to originate.
Chapters
- 0:00From Jane Street to mortgages: Mike Yu joins digital-mortgage startup Blend
- 3:31From Jane Street to mortgages: Mike's Blend role and the shift to online applications
- 6:20From Jane Street to mortgages: Wells Fargo's 3,000 accessibility findings and meeting co-founder Devin
- 10:15What Vesta actually does: replacing 45-day manual mortgage review with an AI-ready system
- 14:06Why rip out the whole system of record: 30,000 flat-file fields block parallel mortgage automation
- 18:41The case for structurally higher rates: Mike Yu cites national debt, AI CapEx, and volatility
- 24:28Is the SaaS apocalypse real? Mike Yu argues vertical SaaS like Vesta can remain valuable
- 30:42Becoming AI-native with Devin: Mike Yu brings rule-triggered agents into Vesta
- 37:46Automation rates and PennyMac's 25%: Vesta estimates 75% progress, but VA access remains a hurdle
- 41:17Why mortgages are so regulated: 30-year fixed loans bring major risk and oversight
- 44:54Why mortgages are so regulated: monthly rule changes and automated underwriting
- 47:04Why mortgages are so regulated: Vesta’s first loan and arduous integrations
- 51:03Why mortgages are so regulated: partner skepticism, customer risk, and hidden know-how
- 53:10Why Vesta survives AGI: lender transformation and slow enterprise adoption
- 56:46Why Vesta survives AGI: mortgage impact, meaningful ownership, and a 95-person remote team
- 59:15Why Vesta survives AGI: 12x revenue growth, hiring, and the enduring craft of product
What the full analysis includes
16 chapters · 91 key points · 16 section summaries · mind map
Preview of the first chapter · From Jane Street to mortgages: Mike Yu joins digital-mortgage startup Blend
The remaining 88 key points, section summaries and mind map are in the full analysis. Analyze this video to see everything →
This video is 1.1 hours. The Free plan analyzes videos up to 30 min; plans from Plus cover videos up to 2 hours.
This is a Tier 1 public summary
Whether the chapter key points, section summaries and mind map are public is up to the person who shared it. Want the full analysis?Submit one yourself.
More from this channel
To Regulate AI Effectively, Focus on How It’s Useda16z Deep DivesMartin Casado argues AI laws should target harmful uses, warning regulatory uncertainty is pushing U.S. startups toward Chinese open-source models.
AI Copilots Are a Dead End. Here's What Actually Works | Kavak CEOa16z Deep DivesKavak CEO Carlos García Ottati says AI agents now handle 90–95% of customer interactions after a year of flat growth.
Are Tariffs Fueling the AI Race? | Deep Dives with a16za16z Deep DivesJacob Helberg says winning the AI race requires leading models, global adoption, secure supply chains, and rebuilding U.S. manufacturing.
Why Modern War Needs Intelligent Power Systems | Chariot Defense CEO on a16za16z Deep DivesChariot Defense’s M424 hybrid system delivered 36 hours of low-signature power as Army leaders push faster procurement and U.S. battery production.
Submarines and the Future of Defense Manufacturing | Hadrian CEO on a16za16z Deep DivesHadrian opened its 2.25-million-square-foot Alabama factory as the Navy faces 70 million submarine-production hours and a severe workforce gap.
Related analyses
What Happens When Everyone Can Build SoftwareY CombinatorSupabase CEO Paul Copplestone says AI builders could drive 1.2 billion personal databases monthly, as Supabase surpasses 13 million users.
EP350. 美國期中選舉預測、英特爾財報大跌、近期台美股市場看法 | M觀點M觀點美國期中選舉共和黨守住參議院機率逾七成,Agentic AI 帶旺 Intel CPU,市場則受資金成本牽動。
AI 時代「大者恆大」,中小企業還剩多少生存空間?|智慧方案執行長 Albert Chen【塞掐 Side Chat】E430INSIDE 硬塞的網路趨勢觀察智慧方案 Albert 用 AI agent 改造紡織接單流程,並警告 AI 導入速度將拉大大廠與小廠差距。
NVIDIA's $2000/Year GPUs and Becoming a Defense ContractorGamers NexusNVIDIA pitches $2,000-per-GPU annual software licenses as its DSX pods expand into defense and battlefield data centers.
Why AI Could Turn Small Law Firms Into Billion-Dollar Businessesa16zEve CEO Jay Madheswaran says AI helped an LA law firm grow from 100 to 500 clients per attorney by automating plaintiff-law work.