Description
Patronus AI is hiring a Researcher to lead foundational research on agentic AI systems, including reinforcement learning, agent simulations, scalable oversight, reward design, behavior analysis, and post-training methods. The role involves owning research projects end to end, designing and evaluating simulation and RL environments, training models with methods such as GRPO and SFT, conducting rigorous experiments, building reproducible production-quality research systems in Python and modern machine learning frameworks, and publishing or open sourcing findings. The position is based full time in Patronus AI’s San Francisco headquarters with five-day in-office attendance, and offers a base salary of $175,000–$300,000 USD plus equity and benefits.
