creators

Pieter Abbeel: Berkeley professor, Amazon frontier models

Pieter Abbeel is a UC Berkeley professor and Amazon Scholar who leads Amazon's frontier model research team and co-founded Covariant and Gradescope.

Pieter Abbeel (born 1977 in Antwerp, Belgium) is a professor of electrical engineering and computer sciences at UC Berkeley, where he directs the Berkeley Robot Learning Lab and co-directs the Berkeley AI Research (BAIR) Lab. He took a KU Leuven MS in 2000 and a Stanford PhD in 2008 as Andrew Ng's first doctoral student, working on apprenticeship learning — including a helicopter taught aerobatics from demonstrations. His deep reinforcement learning work includes trust region policy optimization and domain randomization. He co-founded Covariant, its president and chief scientist from 2017 to 2024, and Gradescope, and hosted The Robot Brains podcast, whose channel stopped posting about three years ago. Amazon licensed Covariant's robotics foundation models and hired its founders in August 2024; he has been an Amazon Scholar since. In December 2025 Amazon told staff he would lead the frontier model research team in its AGI organisation; Amazon Science lists him as co-lead of Frontier AI & Robotics. The professorship continues.

Specialization

robot learning, reinforcement learning, foundation models

Country

United States