RL Engineer: Shape The Future of Humanoid Robots

Apptronik; Austin (TX); On-site; USD 140,000 - 190,000; Full time

Job summary

Apptronik is seeking a Reinforcement Learning Engineer to advance the intelligence and control of our humanoid platforms. You will architect neural topologies, implement advanced RL methods, and develop high-performance policies for whole-body locomotion and manipulation on real hardware.

You will collaborate with hardware and controls teams, contribute to robust training pipelines, and help push safe, scalable deployment of learning-based behaviors on Apollo systems.

Qualifications

• PhD in Computer Science, Robotics or related field, or MS with 2+ years industry experience.
• Proven track record deploying learning-based policies on physical robotic systems.
• Mentoring or technical guidance to other engineers is a plus.
• Strong publication history (CoRL, RSS, ICRA) is a plus.

Responsibilities

• Implement and deploy state-of-the-art RL algorithms for dynamic locomotion and manipulation on hardware.
• Drive the development cycle from prototyping in simulation to transferring policies on the robot.
• Optimize and scale the RL training pipeline for high-throughput simulation and distributed training.
• Develop motion retargeting pipelines to translate human demo data into robust reference trajectories.
• Collaborate with robotics and hardware teams to diagnose issues and enable complex learned behaviors.
• Analyze hardware results and guide future technical directions.

Skills

RL frameworks

Python

C++

Distributed training

RL theory

Robot dynamics

Results-driven

Education

PhD in CS/Robotics

MS + 2+ years industry

Tools

MuJoCo

IsaacGym

PyTorch

JAX

Job description

Apptronik is seeking a Reinforcement Learning Engineer to advance the intelligence and control of our humanoid platforms. You will architect neural topologies, implement advanced RL methods, and develop high-performance policies for whole-body locomotion and manipulation on real hardware.

You will collaborate with hardware and controls teams, contribute to robust training pipelines, and help push safe, scalable deployment of learning-based behaviors on Apollo systems.