Robotics ML Engineer: Real-World Humanoid Policy Dev

Lever, Inc.; Boston (MA); On-site; USD 120,000 - 220,000; Full time

Job summary

FieldAI is seeking a Software/Research Engineer to help build the learning systems that power our next generation of humanoid robots. You will work across whole-body loco-manipulation, reinforcement learning, motion retargeting, and real-world robot deployment, turning new research and technical developments into reliable capabilities on physical humanoids.

This hands-on role closes the loop between research, simulation, and robots operating in the real world, requiring collaboration with

Qualifications

• Experience with RL, imitation learning, or related robotics learning methods.
• Strong robotics fundamentals: kinematics, dynamics, control, physical interaction.
• Experience simulating and testing robotic systems.
• Ability to bridge research and production engineering.
• BS/MS/PhD in Robotics, CS, ML, Engineering, or related field.

Responsibilities

• Develop and train whole-body loco-manipulation policies for humanoid robots.
• Deploy trained policies to physical humanoids and integrate them into production software stack.
• Advance motion retargeting pipeline transforming human motion into robot-executable behaviors.
• Reduce the sim-to-real gap by improving simulation fidelity and real-world calibration.
• Build automated evaluation and validation systems for rapid policy deployment from simulation to hardware.
• Improve the performance and scalability of robot-learning infrastructure for faster experimentation.
• Collaborate with researchers and engineers across perception, learning, simulation, and hardware.

Skills

Reinforcement learning

Imitation learning

Generative models

Robotics fundamentals

Production-grade engineering

Education

BS/MS/PhD in Robotics, CS, ML, Engineering or related

Tools

NVIDIA Isaac Sim

Isaac Lab

Newton

Job description

FieldAI is seeking a Software/Research Engineer to help build the learning systems that power our next generation of humanoid robots. You will work across whole-body loco-manipulation, reinforcement learning, motion retargeting, and real-world robot deployment, turning new research and technical developments into reliable capabilities on physical humanoids.

This hands-on role closes the loop between research, simulation, and robots operating in the real world, requiring collaboration with