Reinforcement Learning Engineer - Locomanipulation

About this role

Humanoid is seeking a Senior or Staff Reinforcement Learning Engineer to develop learning-based control policies for humanoid robots, focusing on dynamic locomotion and manipulation tasks. You'll design training pipelines, optimize sim-to-real transfer, and deploy policies on real robotic systems in an industrial setting.

What you'll do

• Design and train reinforcement learning policies for humanoid robot locomotion and manipulation
• Build scalable simulation and training pipelines using tools like Isaac Lab and MuJoCo
• Create reward functions, observation spaces, and curriculum learning strategies for complex behaviors
• Improve robustness and sim-to-real transfer through iterative simulation and hardware testing
• Deploy and evaluate learned policies on real robotic systems
• Integrate policies into the robot control stack with control and robotics engineers

What they're looking for

• Reinforcement learning algorithms (PPO, SAC, offline RL)
• Robotics applications and real robot deployment
• Physics-based simulation environments (Isaac Lab, MuJoCo)
• Python and/or C++ programming
• Sim-to-real transfer techniques
• Robot dynamics and control
• Whole-body control systems
• Legged robot locomotion

Benefits

• Fully paid medical, dental, and vision insurance with virtual care
• 23 days of PTO plus separate sick leave and paid holidays
• 401(k) retirement plan with employer match
• Equity compensation
• Free daily catered lunch, snacks, and drinks
• Collaboration with top AI and robotics engineers