Software Engineer - Dexterous Manipulation

Apptronik • Austin, TX

1d

About The Position

The Software Engineer – Dexterous Manipulation is a core contributor to the ability of Apptronik’s robots to interact with the world with human-like precision. The role implements, tunes, and deploys reinforcement learning and other learning-based control for high-DOF, multi-fingered robotic hands, translating research in reinforcement and imitation learning and teleoperation retargeting into reliable, production-grade software for simulation and physical hardware.

Working across simulation, hardware, and systems teams, the engineer develops robust manipulation capabilities on real robots. The role supports Apptronik’s human-centered robotics work, including its flagship humanoid robot, Apollo.

Requirements

• Experience with dexterous manipulation, including multi-fingered grasping, in-hand manipulation, and high-DOF hand control.
• Experience applying reinforcement learning to robotic control, including reward design, training, and debugging.
• Strong Python skills and working knowledge of C++ for real-time robotic software.
• Hands-on experience training and validating policies in a physics simulator such as IsaacSim, MuJoCo, or Drake.
• Robotics fundamentals, including kinematics, dynamics, and Jacobian-based control.
• A track record of getting algorithms working on physical hardware, not simulation alone.
• BS, MS, or PhD in Robotics, Computer Science, Electrical Engineering, or a related field.
• At least 3 years of relevant experience in robotic manipulation or complex motion control; recent PhD graduates are considered.
• Evidence of taking complex algorithms from research or simulation to successful deployment on physical hardware.

Nice To Haves

• Imitation learning and diffusion policies.
• Teleoperation using VR or haptics and retargeting.
• Tactile-sensing integration.
• Computer vision, including 6D pose estimation and point clouds.
• End-effector bring-up and calibration.

Responsibilities

• Implement and tune control algorithms for multi-fingered hands, including grasping, in-hand manipulation, and tactile-feedback integration.
• Develop and maintain manipulation software that executes reliably with low latency in the real-time controls stack.
• Translate reinforcement learning and imitation policies, as well as human-to-robot motion retargeting methods, into production-grade C++ and Python.
• Build and refine sim-to-real pipelines, including hand models, domain randomization, and validation in IsaacSim, MuJoCo, or Drake.
• Deploy and debug manipulation capabilities on physical robots, diagnosing sensor-noise, latency, and calibration issues.
• Collaborate with hardware and systems engineers and provide feedback on hand performance and sensor and actuator requirements.
• Uphold code quality through rigorous testing, documentation, and peer review.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level