Reinforcement Learning Engineer - Locomanipulation
Core
Design and train reinforcement learning policies for dynamic locomotion and loco-manipulation behaviors on real humanoid robots.
Role type
Senior/Staff Reinforcement Learning Engineer (Robotics)
Builds
Scalable simulation and training pipelines for sim-to-real transfer of control policies
Domain
Robotics, Reinforcement Learning, Humanoid Robots
Deliverable
production ML models
Required skills
Reinforcement Learning (PPO, SAC, offline RL), Physics-based simulation (Isaac Lab, MuJoCo), Python, C++, Real robotic system deployment, Reward function design, Sim-to-real transfer
Preferred skills
RL for locomotion/legged robots, Robot dynamics, Whole-body control
Responsibilities
Design and train RL policies for humanoid robot control, Build scalable simulation and training pipelines, Design reward functions and observation spaces, Improve robustness and sim-to-real transfer, Deploy and evaluate policies on real robotic systems, Integrate policies into the control stack
Seniority
Senior/Staff, hands-on IC
