[Why I Left Affirm to Build X]
On the decision to leave a good job and start building something from scratch.
About
I'm a machine learning engineer who spent 6 years at Affirm building production ML systems — from a horizontally scalable online feature store to a loan profitability underwriting system that generated $8M+ in contribution margin. I also hold a patent on a simulation-based loan pricing optimization engine.
I studied Data Science and Molecular & Cell Biology at UC Berkeley, a combination that still shapes how I think about modeling: rigorously, from first principles.
Outside of work I shoot film. There's something about the deliberateness of analog photography — one frame at a time, no instant feedback — that I find grounding.
Building
I'm in the early stages of building in ethical AI — my lifelong obsession. I'm not ready to name what it is yet, but I know the space deeply and I'm talking to people who are living the problems I want to solve.
If you're working on something where trust, fairness, or accountability in AI keeps you up at night — in hiring, lending, healthcare, or anywhere decisions matter — I'd genuinely love to hear about it. No pitch, just a conversation.
Tell me about your problem →Photography
[A line about what you shoot — subjects, formats, favorite film stocks.]
Photos coming soon.
Writing
Occasional thoughts on ML, building, and life.