Staff Reinforcement Learning Research Engineer

About the role

Do you want to build the scalable reinforcement learning framework that powers the next generation of humanoid and quadruped robots? As a Staff RL Research Engineer, you'll own the RL stack, including massively parallel simulation, domain randomization, policy optimization, and on-robot deployment. Your job is to make the pipeline fast, reliable, and reproducible. You'll work alongside world-class engineers and scientists pushing the boundaries of whole-body control and dexterous manipulation.

Responsibilities

• Implement on-policy and off-policy learning algorithms
• Scale GPU-accelerated simulation to generate millions of samples per second
• Crack sim-to-real to produce policies that transfer to the physical robot
• Integrate RL with VLAs to fine-tune and distill large multimodal policies
• Make deployment easy, fast, and reproducible
• Build visualization tools that enable data-driven research

Requirements

• MS with 3+ years of experience, or PhD, in ML, Robotics, or a related field
• Deployed policies on physical robots with attention to latency, robustness, and safety
• Expertise with RL toolboxes (RSL-RL, CleanRL, RLlib, Stable Baselines)
• Expertise with simulation and rendering tooling (Isaac Lab, MuJoCo, MjWarp, MjLab)
• Proficient in PyTorch and/or JAX, plus inference runtimes (ONNX, Triton, TensorRT)
• Solid software fundamentals: Bazel, monorepos, Docker, CI/CD

Qualifications

• Deep knowledge of GPU-accelerated physics simulation
• Applied RL to humanoid locomotion, whole-body control, or dexterous manipulation
• Worked on sim-to-real transfer, domain randomization, or system identification
• Experience with heterogeneous compute clusters and Kubernetes

Skills

• Production-grade RL training pipelines

Benefits

• Medical
• Dental
• Vision
• 401(k)
• Paid time off
• An annual bonus structure

Pay

• The base pay range for this position is between $155,284.34- $200,000.

Schedule

• Not specified