Research Engineer / Scientist, Reinforcement Learning

Highlights

• Own post-training for open-weight models built for legal work.
• Pursue research on long-horizon legal reasoning and agentic workflows.
• Take models from experiments through training to customer use.

What they offer

• Equity
• Bonus

Ideal candidate

Hands-on researcher who has owned training runs end-to-end and wants work to reach customers quickly. Thrives in small, high-bar teams, moves fast, and cares about rigorous, mission-driven research.

Requirements

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