San Francisco

Ohm Patel

My current focus is physical reinforcement-learning environments for scientific discovery. My background is in LLM post-training, data quality, and inference.

Previously at Hooglee, an Eric Schmidt-backed AI video startup, and Microsoft AI.

2025–26

HoogleeMember of Technical Staff

LLM post-training and video evaluation.

2025

Microsoft AIResearch Intern

Disaggregated LLM inference simulation.

2024–25

VigilAIFounder

LLM-powered police reports. Backed by Fusen Capital.

B.S. Computer Science, UCF ’25.

ProteoBench

A benchmark for agents analyzing proteomics data.

Amadi

Custom social feeds.

Spydr

1K+ users · Product Hunt daily top 10.

RL environment atlas190 environment types, with sources

A Mechanistic FallibilistOn agents learning through real experiments