Harshit Sikchi
Researcher at OpenAI
San Francisco, California, United States
Summary
👤
Senior🎓
Top SchoolHarshit Sikchi is a research scientist at OpenAI with a decade of experience building and researching reinforcement learning and robot learning systems. He holds a PhD from UT Austin and an MS from Carnegie Mellon, and has a track record of internships and research roles across Meta, NVIDIA, ETH Zurich, and industry startups where he tackled unsupervised RL, multi-agent logistics, and fast simulation. His work spans reward learning, safe and model-based RL, and practical motion-planning solutions—skills honed from leading perception and planning efforts at IIT Kharagpur to developing FlowPlan for low-cost trajectory sampling. Notably, he led a scalable, near-real-time multi-agent logistics solver that made NP-hard problems tractable in large fleets, showing a blend of algorithmic rigor and systems engineering. Based in San Francisco, he focuses on getting robots to learn from limited, heterogeneous data sources and to reason under uncertainty. Outside core research, he combines academic depth with product-minded implementations that push RL methods toward real-world deployment.
10 years of coding experience
9 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at The University of Texas at AustinHigher Secondary Education Science, Higher Secondary Education Science at DAV Public school, KotaMasters Computer Science, Masters Computer Science at Carnegie Mellon UniversityBachelor’s Degree Computer Science, Bachelor’s Degree Computer Science at Indian Institute of Technology, Kharagpur
English, Hindi, Marathi, Sanskrit, Marwari
Github Skills (55)
algorithms9
actor-critic9
convolution8
julia8
maxent8
reinforcement-learning8
image-processing7
robotics7
freenode6
robot6
openshift6
segmentation6
gaussian-processes6
algorithm6
simulation6
Programming languages (6)
Github contributions (5)
hari-sikchi/pytorch_CPO
Jul 2021 - Aug 2022
Simple implementation for Constrained Policy Optimization in Pytorch
Contributions:1 release, 9 commits, 8 pushes in 1 year 1 month
pytorch
hari-sikchi/offline_rl
Aug 2020 - Aug 2022
Pytorch implementation of state-of-the-art offline reinforcement learning algorithms.
Contributions:1 release, 22 commits, 8 pushes in 2 years
pytorchreinforcement-learningoffline-reinforcementoffline-rl
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.