Harsh Girase

Harsh Girase is a Machine Learning Engineer at Apple since April 2025, specializing in post-training processes. Prior to this role, Harsh served as a Senior Machine Learning Engineer at Cruise from January 2022 to April 2025, focusing on foundational models for autonomous vehicle planning. Harsh's experience includes significant research contributions at Berkeley Artificial Intelligence Research from January 2018 to December 2023, involving trajectory prediction and video anticipation, with publications at ECCV, CVPR, and ICCV, as well as work at the UC Berkeley HART Lab on motion capture classification. Previously, Harsh was a Machine Learning Researcher at Honda Research Institute USA from January 2021 to May 2022, where projects included the LOKI model for trajectory prediction. Harsh also completed a prediction and perception internship at Uber in mid-2020, concentrating on behavior prediction for self-driving cars. Educational qualifications include a Master of Science in Electrical Engineering and Computer Sciences and a Bachelor’s degree in Computer Science, both from the University of California, Berkeley.