About the job
In this role, you'll combine hands-on technical work with leadership, ensuring your team delivers the best environment to develop robust physical AI solutions for dynamic real-world environments. You'll leverage Amazon's vast computational resources to tackle ambitious problems in areas like very large multi-modal robotic foundation models and efficient, promptable model architectures that can scale across diverse robotic applications.
Responsibilities
You will lead the ML infrastructure team to create model training and simulation environment for developing large robotics foundational models for reasoning, perception, locomotion, and manipulation.
Work together with AI researchers to implement and optimize training of new model architectures at a large scale
Define roadmap and build physics realistic simulation environment for reinforcement learning, closed-loop simulations and synthetic data generation.
Implement tooling for data creation, model experimentation, and continuous integration
Qualifications
Minimum
10+ years of engineering experience
5+ years of engineering team management experience
Experience managing multiple concurrent programs, projects and development teams in an Agile environment
Proficiency in Python and C++, experience in building large-scale training infrastructure using PyTorch
Experience with Isaac Sim, Unity, Unreal or proprietary 3D game engine, or industry-equivalent technology (3D animation, simulation, etc)
10+ years of planning, designing, developing and delivering infrastructure software
Experience partnering with product or program management science teams working in research environments
Preferred
Experience designing and developing large scale, high-traffic applications
Experience with physical robots, reinforcement learning, synthetic data generation. Experience optimizing physics simulation for articulated robots and rigid body interactions.
Deep understanding of different model architecture like VLM, Imitation learning, VLA
Experience implementing techniques from research papers