Pieter Abbeel — Pioneer of Deep Reinforcement Learning for Robotics

Home  /  AI Inventors  /  Pieter Abbeel
Pieter Abbeel is a UC Berkeley professor and one of the most influential researchers in robot learning, whose work on deep reinforcement learning underpins much of modern robotics AI. He has co-founded two companies — Gradescope and Covariant — and now co-leads Amazon's Frontier AI & Robotics team.

Pieter Abbeel was born in Antwerp, Belgium, in 1977. He earned his bachelor's and master's degrees in electrical engineering at KU Leuven before moving to the United States for his PhD in computer science at Stanford University, where he studied under AI researcher Andrew Ng. His 2008 doctoral thesis on apprenticeship learning and reinforcement learning for robotic control became foundational to a research career built on a simple idea: robots should be able to learn skills the way humans do, by trial, error, and demonstration, rather than through hand-coded rules.


Abbeel joined UC Berkeley as a professor in 2008, where he now directs the Berkeley Robot Learning Lab and co-directs the Berkeley Artificial Intelligence Research (BAIR) Lab. His lab's research has enabled robots to perform advanced helicopter aerobatics, tie knots, fold laundry, and learn locomotion and manipulation skills through deep reinforcement learning — work that has produced widely used techniques including TRPO, SAC, and domain randomization, alongside contributions to diffusion models now central to image and video generation. He has received numerous honors for this work, including the Presidential Early Career Award for Scientists and Engineers (PECASE), an NSF CAREER Award, MIT Technology Review's TR35, IEEE Fellow status, and the ACM Prize in Computing.


Alongside his academic work, Abbeel has co-founded two companies. In 2014, he co-founded Gradescope, an AI-assisted grading platform later adopted by more than 500 universities and acquired by Turnitin in 2018. In 2017, following a stint as a research scientist at OpenAI, he co-founded Covariant with three of his students, building foundation models that let robotic arms handle warehouse tasks like bin-picking. In August 2024, Amazon hired Covariant's three founders — Abbeel, Peter Chen, and Rocky Duan — along with roughly a quarter of the company's staff, while signing a non-exclusive license to use Covariant's AI models; Covariant itself continued to operate independently. Abbeel now co-leads Amazon's Frontier AI & Robotics team as an Amazon Distinguished Scientist, alongside his continuing Berkeley professorship. He is also an investing partner at AIX Ventures and hosts The Robot Brains Podcast, interviewing leading figures in AI and robotics. Perhaps uniquely among robotics researchers, his former students have gone on to co-found more than a dozen AI companies, including OpenAI, Perplexity, Physical Intelligence, and Skild AI.

Pieter Abbeel
Pieter Abbeel
Pieter Abbeel
Company University of California, Berkeley; Amazon (Frontier AI & Robotics); Covariant (technology licensed to Amazon in 2024)
Country Belgium (born and raised in Belgium; based in the United States since his PhD)
Born February 04, 1977 (age 49)
Antwerp, Belgium
Education KU Leuven, Belgium — BS/MS, Electrical Engineering (2000); Stanford University — PhD, Computer Science (2008), advised by Andrew Ng
Notable work Pioneering deep reinforcement learning for robotics (including TRPO, SAC, domain randomization); apprenticeship-learning-based autonomous helicopter aerobatics; robotic laundry-folding and knot-tying; co-founder of Gradescope and Covariant; host of The Robot Brains Podcast
Title Role: Professor of Electrical Engineering and Computer Sciences, UC Berkeley; Director, Berkeley Robot Learning Lab; Co-Director, Berkeley AI Research (BAIR) Lab; Amazon Distinguished Scientist / Amazon Scholar, co-leading Amazon's Frontier AI & Robotics team; Co-Founder, Covariant; Co-Founder, Gradescope; Investing Partner, AIX Ventures

Share this AI Inventor Biography

Spread historical AI biographies and inspire future innovators.

Share Post
← Previous Biography Grant Lee
Next Biography → JJ Fiasson

Related AI Inventors

Microsoft; formerly Obviously AI, Inc. (Zams); Hampshire College

Nirman Dave

Nirman Dave is an Indian AI entrepreneur and product innovator best known for co-founding Obviously AI with Tapojit Debnath. Their platform helped non-technical business users build predictive machine learning models without writing code. Dave developed the original product concept after working on customer-behavior predictions at Streamlabs. Founded through academic work beginning in 2018 and publicly launched in 2020, Obviously AI later evolved into Zams, an enterprise AI agent platform. Recognized in Forbes' 2022 Enterprise Technology 30 Under 30 list, Dave announced Zams' wind-down in 2026 and subsequently joined Microsoft to work on its Discovery AI for Science platform.

AutoGPT / Determinist Ltd

Toran Bruce Richards

Toran Bruce Richards is a British software developer, entrepreneur, and artificial intelligence innovator best known as the creator of AutoGPT, one of the most influential early open-source autonomous AI agent projects. Introduced on March 30, 2023, AutoGPT demonstrated how large language models such as GPT-4 could combine planning, memory, software tools, and iterative execution to pursue complex user-defined goals.

LlamaIndex; previously Robust Intelligence, Uber Advanced Technologies Group, Quora; Princeton University

Jerry Liu

Jerry Liu is an American-based AI technology entrepreneur, software developer, and machine learning researcher best known as the original creator of GPT Index and co-founder and CEO of LlamaIndex. He graduated summa cum laude from Princeton University in 2017 with a degree in Computer Science and a certificate in Finance.