Pulkit Agrawal

Pulkit Agrawal is an Associate Professor in the Department of Electrical Engineering and Computer Science (EECS) at the Massachusetts Institute of Technology (MIT). He leads the Improbable AI Lab, which is part of the Computer Science and Artificial Intelligence Lab (CSAIL), affiliated with the Laboratory for Information and Decision Systems (LIDS), and involved with the NSF AI Institute for Artificial Intelligence and Fundamental Interactions (IAIFI).

Dr. Agrawal completed his Ph.D. at UC Berkeley and undergraduate studies at IIT Kanpur. He co-founded SafelyYou Inc., which builds fall prevention technology, and serves as an advisor to Tutor Intelligence, Lab0 Inc., and Common Sense Machines. His research focuses on computational sensorimotor learning, aiming to build machines with human-like manipulation and locomotion abilities. His work spans perception, control, hardware design, robotics, and reinforcement learning, with inspiration drawn from cognitive science and neuroscience.

His recent publications demonstrate a strong trend toward integrating foundation models with robotics, developing methods for sim-to-real transfer, and exploring language model personalization for robotic applications. The research spans both theoretical advances in reinforcement learning and practical implementations in robotic systems, with significant contributions to dexterous manipulation, locomotion, and AI for physical systems.

• IIT Kanpur Young Alumnus Award
• 2024 IEEE Early Academic Career Award in Robotics and Automation
• Amazon Research Awards (2020, 2019, 2019, 2023)
• Best Paper Award at Conference on Robot Learning (CoRL) 2021

Dr. Agrawal advises numerous graduate students, postdocs, and undergraduate researchers through his Improbable AI Lab. His lab has received substantial funding from Amazon Research Awards and has mentored students who have gone on to positions at Google, Boston Dynamics, OpenAI, and leading academic institutions. The Improbable AI Lab brings together researchers from computer science, electrical engineering, mechanical engineering, and cognitive science to tackle challenging problems in robot learning and sensorimotor intelligence.