Robotics Software Engineer - Dexterous Manipulation

Apptronik; Austin (TX); On-site; USD 110,000 - 170,000; Full time

Job summary

Apptronik seeks a Software Engineer to advance dexterous manipulation for humanoid robots. You will translate state-of-the-art reinforcement and imitation learning into robust software that runs in simulation and on hardware, tackling high-DOF hand control and low-latency real-time execution across teams.

The role emphasizes building sim-to-real pipelines, validating policies in physics engines, and collaborating with hardware and systems engineers to push Apollo-like capabilities toward

Qualifications

• Dexterous manipulation expertise with multi-fingered grasping and in-hand manipulation.
• Proficient in reinforcement learning for robotic control; experience with imitation learning or diffusion policies a plus.
• Strong Python and C++ for real-time robotic software.
• Hands-on experience training and validating policies in physics simulators (IsaacSim, MuJoCo or Drake).
• Foundations in robotics: kinematics, dynamics, and Jacobian-based control.
• Proven track record deploying algorithms on physical hardware, not just simulation.
• Nice-to-have: teleoperation (VR/haptic), tactile sensing, CV (6D pose/point clouds).

Responsibilities

• Implement and tune control algorithms for multi-fingered hands including grasping and in-hand manipulation.
• Develop and maintain manipulation software with low-latency real-time execution.
• Build and refine sim-to-real pipelines and validation workflows.
• Deploy and debug manipulation capabilities on physical robots; diagnose sensor noise and calibration issues.
• Collaborate with hardware and systems engineers to optimize hand performance and actuators.
• Uphold code quality through testing, documentation, and peer review.

Skills

Multi-finger manipulation

Reinforcement learning

Python and C++

Physics simulators

Kinematics & Jacobian control

Deploy on hardware

Teleoperation / haptics

Education

BS/MS/PhD in Robotics, CS, EE, or related

3+ years in robotic manipulation / motion control

Experience transferring algorithms to hardware

Tools

IsaacSim

MuJoCo

Drake

Job description

Apptronik seeks a Software Engineer to advance dexterous manipulation for humanoid robots. You will translate state-of-the-art reinforcement and imitation learning into robust software that runs in simulation and on hardware, tackling high-DOF hand control and low-latency real-time execution across teams.

The role emphasizes building sim-to-real pipelines, validating policies in physics engines, and collaborating with hardware and systems engineers to push Apollo-like capabilities toward