Pieter Abbeel

Pieter Abbeel is a Belgian-born AI researcher and a professor at UC Berkeley, where he directs the Robot Learning Lab and co-directs the Berkeley AI Research (BAIR) Lab. A leading figure in reinforcement learning and robot learning, he completed his PhD at Stanford under Andrew Ng, joined OpenAI as a researcher in 2016, co-founded the grading startup Gradescope (acquired by Turnitin), and in 2017 co-founded the robotics-AI company Covariant. He also hosts “The Robot Brains” podcast.

After Amazon licensed Covariant's robotics foundation models in 2024, Abbeel joined Amazon, where he helps lead large-language-model efforts while retaining his Berkeley professorship. Few academics have trained as many of today's AI founders and leaders. A generation of his students and collaborators have gone on to found or lead prominent AI labs and startups.

Abbeel's ties to frontier-AI cap tables run unusually deep: many founders are his former Berkeley students, including Perplexity CEO Aravind Srinivas. He is an Investment Partner at the AI-focused firm AIX Ventures, and Dealroom tracks his investments in Perplexity, Cohere, Waabi and Workera. His backing concentrates in robotics, foundation models and the applied-AI companies closest to his research.