Robotics Foundation Models Engineer (human)

The role

This person will drive the development of cognitive robotics applications using Physical AI and the NeuraGym platform. They will execute the full pipeline from foundation models to production-ready robotic intelligence and deploy policies to real-time control.

What you will do

• Drive the development of cognitive robotics applications through Physical AI and the NeuraGym platform
• Execute the complete NeuraGym pipeline including teleoperation, data annotation, training, simulation, deployment, and validation
• Define the data strategy and own the full post-training loop
• Deploy policies from checkpoint to real-time closed-loop control
• Work across NEURA OS and the control layer
• Perform hands-on system integration on MAiRA, MiPA, LARA, and humanoid platforms
• Guide and onboard users to NeuraGym
• Serve as a key interface between Cloud and AI teams
• Create evaluation harnesses, success-rate benchmarks, and regression tests
• Prototype techniques from recent papers and bring them into production deployments

Required qualifications

• Master's degree or PhD in Computer Science, Robotics, Electrical Engineering, or a related field
• Hands-on machine-learning experience in robotics in at least one of vision-based manipulation, reinforcement/imitation learning, or multimodal models
• Experience fine-tuning large pretrained models for robotics or another domain including PEFT/LoRA, full fine-tunes, dataset curation, distillation, and evaluation
• Solid Python skills
• Practical experience with PyTorch or TensorFlow
• Working command of a robotics software stack including robot operating systems and middleware, real-time control, kinematics and motion planning, and force/impedance control
• Solid understanding of robot hardware including drives, joint modules, torque sensing, encoders, end-effectors, sensor integration, calibration, electrical and fieldbus interfaces, and functional safe
• Proven problem-solving abilities
• Ability to handle multiple projects in parallel
• Solid command of English and German (B2–C1 level)
• Willingness to travel to client sites and other NEURAGym locations

Nice to have

• Familiarity with the open VLA and policy-learning ecosystem (e.g., π0/π0.5, OpenVLA, LeRobot, diffusion policies)
• C++
• AWS, Azure, or GCP experience
• Familiarity with robotic simulation tools (IsaacSim, MuJoCo, etc.)

About the company

neura-robotics-gmbh