Utkarsh Patel
Senior Software Engineer - Machine Learning at Google
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolUtkarsh Patel is a Senior Software Engineer specializing in machine learning, computer vision, and robotics with nine years of experience building production-grade ML systems at Torc Robotics and Google. He has tackled real-time perception for self-driving trucks and now applies ML at scale on Google Earth AI and sustainability initiatives. His background blends academic research—developing an open-source reinforcement learning toolkit (garage) and robotics systems—with hands-on engineering like converting PyTorch models to TensorRT and implementing active learning and calibration for object detection. Comfortable across TensorFlow and PyTorch, he brings both research rigor and pragmatic optimization to deployable pipelines. Based in New York, he combines a USC MS in Intelligent Robotics with a keen interest in reproducible RL and real-world autonomous systems.
9 years of coding experience
9 years of employment as a software developer
Master of Science - MS Computer Science (Intelligent Robotics), Master of Science - MS Computer Science (Intelligent Robotics) at University of Southern CaliforniaBachelor of Science (B.Sc.) Computer Science, Bachelor of Science (B.Sc.) Computer Science at Cleveland State University
Hindi, Gujarati, English
Github Skills (13)
algorithm10
mle10
algorithms10
pytorch10
rep10
machine-learning10
tensorflow10
repr10
python10
ml10
ppp9
gymnasium9
openai-gym9
Programming languages (3)
Github contributions (5)
rlworkgroup/garage
Oct 2018 - Nov 2019
A toolkit for reproducible reinforcement learning research.
Role in this project:
ML Engineer
ML EngineerContributions:10 commits, 14 PRs, 117 pushes in 1 year 1 month
Contributions summary:Utkarsh made substantial contributions to the `garage` repository, a toolkit for reinforcement learning research. Their work focused on replacing the `CategoricalMLPPolicy` in tests, indicating a focus on algorithm implementation and performance comparison. Additional commits involve adding a multi-environment wrapper and example, and also include a PyTorch implementation of TRPO and related improvements, highlighting their expertise in different ML frameworks (TensorFlow and PyTorch) and reinforcement learning algorithms. The user also addressed testing and optimization concerns.
reinforcement-learningrl-algorithmsreproducibilitypytorchtensorflow
usc599resl/mlgm
Mar 2019 - Jul 2019
Meta learning for generative models.
Contributions:21 commits, 19 pushes, 4 branches in 4 months
generative-modelmeta-learning
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.