Tsubasa Onishi builds physical AI and autonomous agent systems at NVIDIA. He cares about making synthetic data generation practical for robot learning pipelines.
His background combines scientific machine learning and high-performance computing. He teaches roboticists how to generate training datasets using Isaac Sim and manage persistent agent state with the OpenClaw gateway.
Before NVIDIA, he spent time at Chevron, ETH Zurich, and Los Alamos National Lab. He holds a PhD in petroleum engineering from Texas A&M University.