Founding Backend Engineer - AI for Commercial Lending

Want to be a founder in everything but the title, at the moment 8 pilots turn into seven-figure contracts? A YC-backed startup is building the AI operating system for commercial lending. At banks and fintechs, underwriting and servicing a business loan often takes months; their agents bring that down to under two weeks. The agents keep an eye on loan portfolios, ask borrowers for missing documents, read whatever arrives (from crisp PDFs to blurry scans and handwritten notes) and spread the financials according to each lender's own rules, with in-house lending specialists checking every result. It's a $3T market, and a bank spends roughly $25k to originate one loan. What sets them apart: underwriting models tuned per industry, because general-purpose LLMs still miss those nuances. They're hiring 3 founding engineers. What you'll own A production agent from start to finish, probably beginning with the portal borrowers use Computer vision and parsing that read financial statements, and live reconciliation when the numbers disagree The backend and infra that get the product from pilot to scale Time with customers at their offices, and turning those conversations directly into product What you bring 2-6 years as a professional engineer, with real startup experience (non-negotiable) Deep backend and infra skills, and AI, ML or agent systems you've put into production Clear, to-the-point communication; happy to sit with customers and travel monthly Claude, Cursor or Codex as part of your everyday workflow Something in your background that makes people stop and look twice Based in SF or ready to move, and able to start soon Bonus points You founded a company that took off ICPC or olympiad medals Palantir-style forward deployed work, or early engineer at a fast-growing fintech Document AI, computer vision, MCP or agent infrastructure What’s in it for you $200k-$300k base + 1-3% equity Founding-team ownership across product and engineering Visa sponsorship welcome Good to know Full-time, in person in San Francisco, 9-9-6 (Saturdays can be remote) Customer visits roughly once a month Process: intro call -> technical interview -> optional behavioral chat -> work trial in person -> offer #J-18808-Ljbffr Rnb Consultancy