Senior AI Controls Engineer - Humanoid Robotics

Agility Robotics; Pittsburgh (Allegheny County); Hybrid; USD 175,000 - 292,000; Full time

Benefits offered by this job

401k match

Stock options

Insurance coverage

Flexible work

Relocation assistance

Annual bonus

Parental leave

On-site meals

Job summary

Agility Robotics is seeking an AI Controls Engineer to design, train, and deploy high-rate RL policies for humanoid Digit, enabling robust locomotion and manipulation in dynamic environments. You will work on perception-enabled control, integrating sensors to achieve safe, adaptive robot behavior in real-world facilities.

The role emphasizes creating scalable RL pipelines, sim-to-real transfer, and collaboration with on-robot teams to ship production-ready policies for deployment in warehouses

Qualifications

• 4+ years of experience developing and deploying RL policies for robotics.
• Strong Python skills and PyTorch experience.
• Experience designing reward functions and tuning hyperparameters.
• Perception-in-the-loop control with real-time sensory inputs.
• Proven RL deployment on real robots.
• Familiarity with robot simulation environments and sim-to-real transfer.

Responsibilities

• Design, train, and deploy robust RL policies for locomotion, manipulation, and whole-body control.
• Integrate perception with RL for obstacle-aware, collision-free motion.
• Develop and maintain RL infrastructure, including training pipelines and evaluation frameworks.
• Create new simulation environments to support training and evaluation.
• Collaborate with on-robot software and deployment teams to ship production-quality policies.

Skills

RL for robotics

Python

PyTorch

Reinforcement learning

Sim-to-real transfer

Education

Advanced degree (MS/PhD) in Robotics or CS

Tools

Mujoco-Warp

Isaac

Job description

Agility Robotics is seeking an AI Controls Engineer to design, train, and deploy high-rate RL policies for humanoid Digit, enabling robust locomotion and manipulation in dynamic environments. You will work on perception-enabled control, integrating sensors to achieve safe, adaptive robot behavior in real-world facilities.

The role emphasizes creating scalable RL pipelines, sim-to-real transfer, and collaboration with on-robot teams to ship production-ready policies for deployment in warehouses