I build production AI agents and the evals that check them, and I'm learning to train agents with RL. I'm an AI Tech Lead at Coefficient; before that I built backend and distributed systems at Coinbase and Flipkart.

I also found an explicit family of counterexamples to the Jacobian conjecture in every dimension n ≥ 3, later discussed on Terence Tao's blog. Say hi on X or by email.

Now: Building autograd from scratch, week 2 of learning post-training.