Overview In this role you will advance humanoid robot foundation models and embodied agents within NVIDIA's GEAR group. You will shape scalable AI for both virtual and physical environments, collaborating with cross-disciplinary teams to transfer research into products. Your work will push multimodal models, large-scale training, and robotics control to enable autonomous, capable agents. This is a moonshot role with potential to influence major projects and product roadmaps.
Compensation / Benefits- equity
- benefits
- competitive base salary
- opportunity to impact moonshot research projects
- lab access and hardware for robotics
Responsibilities- Design and implement novel AI algorithms for humanoid robots and embodied agents
- Develop large-scale training and inference methods for foundation models
- Optimize deployment of AI models in simulation and on robot hardware
- Collaborate with research and engineering teams across NVIDIA to transfer research to products and services
Key requirements- PhD in Computer Science/Engineering or equivalent research experience
- 5+ years of work/research experience in Multimodal Foundation Models or Hands-on robotics learning
- Publications or demonstrated expertise in LLMs, large vision-language models, video diffusion, or action-based transformers
- Strong programming and ML engineering skills (Python, PyTorch, Jax, TensorFlow) with C++ and CUDA a plus
- Experience with large-scale ML systems and compute infrastructure
- Hands-on robotics experience: reinforcement/imitation learning, control methods; strong Python, C++, ROS
- collaborative
- strong communication
- problem-solving
- PyTorch
- Jax
- TensorFlow
