AI Training Infrastructure Engineer – Humanoid Whole Body Control

Figure • San Jose, CA

2d • Onsite

About The Position

Figure is an AI Robotics company focused on autonomous general-purpose humanoid robots, with the goal of shipping robots with human-level intelligence for home and commercial markets. The company is based in North San Jose, CA, and requires 5 days/week in-office collaboration. This role is for an engineer to own the training and deployment infrastructure for RL-based whole-body control systems. It combines robotics, machine learning, controls, and software systems engineering, and is crucial for accelerating the iteration, training, and deployment of new capabilities to the fleet of humanoid robots.

Requirements

• Strong software engineering fundamentals with production experience in Python and PyTorch
• Experience building or scaling training infrastructure for robotics, control systems, or large-scale ML workloads
• Familiarity with physics simulation tools such as NVIDIA PhysX, MuJoCo, Warp, or PyBullet
• Working knowledge of dynamics, controls, and robotics systems
• Experience with reinforcement learning, imitation learning, or policy distillation
• Strong ownership mindset—you own systems that your teammates rely on every day
• Experience modeling contact interactions and photorealistic simulation environments for complex manipulation tasks

Nice To Haves

• Experience with humanoid or legged robot control
• Background in distributed systems, job schedulers, or cluster management
• Experience deploying ML models or control policies to real-world systems

Responsibilities

• Own and scale the infrastructure used to train whole-body control policies (simulation, data pipelines, orchestration, visualizations)
• Design systems that are fast, reliable, and highly configurable for our controls engineers
• Ensure high cluster utilization and minimal downtime—unblocking the team and accelerating iteration cycles
• Evaluate and integrate physics engines, simulation environments, and parameterizations to balance realism and training speed
• Optimize hyperparameters and infrastructure to maximize training speed and efficiency and final model performance
• Build robust tooling to take policies from training → validation → deployment on hardware

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: No Education Listed