Software Engineer - RL Environments

What you'll do

• Construct simulated environments and data that expose model failure modes; analyze agent trajectories and run experiments.
• Build evaluation rubrics and reward signals for RLHF and RLVR pipelines, and create frameworks to measure dataset quality and impact.
• Manage real-world and synthetic data pipelines and partner with research teams on training objectives, evaluations, and infrastructure.

What they ask for

• Ability to design lightweight experiments, move quickly, and extract actionable insights from messy results.
• Experience using Docker or similar containerization tools to design and monitor systems at scale.
• Strong familiarity with common reinforcement learning algorithms and methods, especially for post-training LLMs.

Nice to have

• Experience at an RL environment company or an AI safety or benchmarking organization such as METR or Artificial Analysis.
• Experience as a founder or early engineer at an early-stage startup.

Pay

$260K – $290K USD base salary; $200K – $400K USD bonus

Tools

Docker, RLHF, RLVR

AI focus

Develops environments, data, reward signals, and evaluations for reinforcement learning and post-training of frontier models; experiments inform model training, alignment, and capability.