Senior AI Software Engineer, AI Controls
Core
Design, train, and deploy reinforcement learning (RL) policies for humanoid locomotion, whole-body control, and manipulation to enable safe, perceptive motion in dynamic real-world environments.
Role type
Senior AI Controls Engineer (Humanoid Robotics)
Builds
High-rate learned controllers and RL policies for the Digit humanoid robot
Domain
Robotics, Reinforcement Learning, Humanoid Locomotion
Deliverable
production ML models
Required skills
Reinforcement Learning (RL) policy development, Python, Deep Learning (PyTorch), Reward function design, Hyperparameter tuning, Perception-in-the-loop control, Sim-to-real transfer, Robot simulation (Mujoco-Warp, Isaac)
Preferred skills
Contact-rich manipulation, Policy distillation, Top-tier ML/Robotics conference publications
Technologies
PyTorch, Mujoco-Warp, Isaac Gym/Sim
Responsibilities
Design and train robust RL policies for locomotion and manipulation; Integrate perception into RL policies for collision-free motion; Develop and maintain scalable RL training pipelines and evaluation frameworks; Design simulation environments for policy training; Collaborate with deployment teams to ship production-quality policies
Seniority
Senior, hands-on IC
