Research Engineer / Scientist, Reinforcement Learning

Highlights

• Own post-training recipe for open-weight models built for legal work.
• Improve long-horizon legal reasoning and agentic workflows.
• Build rigorous evals to decide what ships to customers quickly.

What they offer

• equity
• bonus

Ideal candidate

Hands-on researcher who has owned training runs at scale and wants work in users' hands fast. Thrives in small high-bar teams, takes end-to-end ownership, and is driven to crack legal reasoning.

Requirements

• Hands-on experience post-training LLMs with SFT and RL including preference optimization
• Direct ownership of training runs and results
• Deep ML fundamentals and reliable distributed training experience
• Strong Python engineering with production-quality code and end-to-end ownership
• Research experience in RL for language models or reward modeling