AI Training Infrastructure Engineer – Humanoid Whole Body Control
Core
Building and scaling the training and deployment infrastructure for RL-based whole-body control systems in autonomous humanoid robots.
Role type
Senior IC infrastructure engineer (robotics/ML)
Builds
Simulation environments, data pipelines, orchestration systems, and tooling for policy deployment to hardware
Domain
Robotics, machine learning, controls, and software systems engineering
Deliverable
infrastructure
Required skills
Python, PyTorch, reinforcement learning, imitation learning, policy distillation, physics simulation (NVIDIA PhysX, MuJoCo, Warp, PyBullet), dynamics, controls, robotics systems, distributed systems, job schedulers, cluster management
Preferred skills
Humanoid or legged robot control, deploying ML models to real-world systems
Technologies
Python, PyTorch, NVIDIA PhysX, MuJoCo, Warp, PyBullet
Responsibilities
Own and scale infrastructure for training whole-body control policies; design fast, reliable, configurable systems for controls engineers; ensure high cluster utilization and minimal downtime; evaluate and integrate physics engines and simulation environments; optimize hyperparameters and infrastructure for training speed and efficiency; build tooling to move policies from training to validation to hardware deployment
Seniority
Senior, hands-on IC
