Staff Software Engineer - Robotic Simulation

Apptronik Apptronik · Robotics · HQ · Software Engineering

Staff Software Engineer to lead the architecture and development of core simulation infrastructure for a human-centered robotics company. This role will focus on building high-performance digital twins for Reinforcement Learning (RL) and Vision-Language-Action (VLA) model training, controls validation, and CI/CD integration testing, with a strong emphasis on sim-to-real transfer and photorealistic perception.

What you'd actually do

  1. Own the technical roadmap and architecture of the core simulation platform, building high-performance digital twins that serve as the foundation for Reinforcement Learning (RL) and Vision-Language-Action (VLA) model training, controls validation, and CI/CD integration testing.
  2. Design and scale cloud-native simulation pipelines capable of generating millions of experience hours per day, leveraging distributed computing frameworks to parallelize physics and rendering for rapid policy iteration.
  3. Bridge the reality gap by developing advanced contact models and actuator dynamics, ensuring that policies trained in simulation transfer zero-shot to physical humanoid hardware.
  4. Lead the development of sensor-realistic environments (camera, LiDAR, depth) that challenge the perception stack, creating the dynamic and diverse worlds necessary to robustify robot vision capabilities.
  5. Serve as the technical authority on simulation, collaborating with Control and AI research teams to anticipate future needs and drive the evolution of tools.

Skills

Required

  • C++
  • Python
  • CI/CD
  • automated testing
  • code quality standards
  • technical leadership
  • robotics concepts (kinematics, dynamics, controls, path planning, system identification)
  • modern robotic simulators (IsaacLab, MuJoCo)
  • architecting and managing large-scale training workloads
  • deploying parallel simulations on cloud platforms (AWS, GCP, Azure)
  • distributed computing frameworks (Ray, Kubernetes)

Nice to have

  • Reinforcement Learning
  • VLAs
  • game engines (Unity, Unreal)

What the JD emphasized

  • reinforcement learning
  • Vision-Language-Action (VLA)
  • humanoid hardware
  • robot vision capabilities

Other signals

  • simulation infrastructure
  • reinforcement learning
  • humanoid robot
  • embodied AI