Commercial credit is a $3+ trillion market in the US — and almost none of it has been automated.
There are 33 million small and mid-sized businesses in the United States. Almost all of them run on credit. And the way banks decide who gets it hasn't meaningfully changed since the 1980s — a spreadsheet, a stack of PDFs, a 60-day wait, and a decision that often comes back "no" for reasons nobody can articulate. Banks are running on bloated operations teams, losing deals and working with antiquated software that works half the time.
We're building autonomous lending agents that fix this — the infrastructure for how commercial credit gets issued, and how our economy grows.
Proximitty (YC W26) is building autonomous commercial lending teams for the largest banks and fintechs in the US. We're backed by top investors — Y Combinator, CRV, Cohen Circle, and Uphonest, with angels leading teams at Stripe and OpenAI.
Our founding team:
Wye Yew (CEO) advised banks at McKinsey on commercial lending and credit risk and previously led Growth at Taptap Send ($75M -> $200M ARR in 12 months). He was the first Malaysian to place in the Top 3 in the World Mathematics Championships in 2018.
Zi (CTO) built the world's first MCP gateway at ACI.dev (4k+ Github stars in 3 weeks), and prior to that led Bloomberg's core security infrastructure, which powers 300K+ terminals today. He previously received the National Scholarship of China.
You’ll be working alongside founders who’ve led fintechs to Series D and shipped at scale
on agent infrastructure problems most teams haven’t earned the chance to solve yet — self-evolving & self-healing agents, software factory, and large scale evals. If we get this right, millions of businesses currently ignored by banks get access to the credit they should already have.
As our first founding engineer, you’ll work directly with the founders on frontier applied-AI problems: long-horizon reasoning over years of complex financial data, reliable computer-use agents in brittle legacy systems, and evaluation infrastructure for high-stakes decisions where correctness is partly factual, partly judgment. In your first 6–12 months, you will: