Samir Patil

ml engineer · pune, india

I work on online RL for tool-using agents. Founding engineer at RunWhen (building agentic AI for autonomous SRE); previously L5 ML at Google Maps Ads.

I write here about training systems, post-training, and what I learn from the work. Find me on X, GitHub, HuggingFace.


Writing

First post landing shortly.

all writing →