Sarthak Chakraborty
Graduate Research Assistant
Champaign, Illinois, United States
Summary
👤
Senior🎓
Top SchoolSarthak Chakraborty is a PhD candidate at UIUC specializing in ML-for-systems, distributed systems, and cloud reliability, with nine years of experience spanning industry research and academic projects. He has driven federated learning and scalable ML deployment work at Adobe—building asynchronous FL frameworks that ran on mobiles, browsers, Raspberry Pi and virtual desktops—and interned on cloud system reliability at Microsoft. At UIUC he contributes to the Distributed Protocols Research Group and is an Illinois Distinguished Fellow, while maintaining open-source contributions such as enhancements to TensorFlow examples for TFLite transfer learning and Keras Functional API support. He holds a dual B.Tech+M.Tech from IIT Kharagpur with top honors and brings a blend of rigorous theory (EM, ALS/CP tensor methods) and practical systems engineering to production-ready ML systems. Notably, his work demonstrates an uncommon focus on making cutting-edge ML algorithms portable across diverse client runtimes and on improving model conversion and gradient handling in widely used tooling.
9 years of coding experience
2 years of employment as a software developer
High School, Science, ISC, High School, Science, ISC at St. Joseph's College, KolkataDoctor of Philosophy - PhD, Computer Science, 4.0 / 4.0, Doctor of Philosophy - PhD, Computer Science, 4.0 / 4.0 at University of Illinois Urbana-ChampaignM. Tech 5yr dual degree, Computer Science and Engineering, 9.74, M. Tech 5yr dual degree, Computer Science and Engineering, 9.74 at Indian Institute of Technology, Kharagpur
English, Hindi, Bengali
Github Skills (6)
keras10
machine-learning10
tensorflow-lite10
tensorflow10
python9
transfer-learning9
Programming languages (7)
Github contributions (5)
tensorflow/examples
Jun 2020 - Aug 2020
TensorFlow examples
Role in this project:
ML Engineer
ML EngineerContributions:6 commits, 2 PRs, 4 comments in 2 months
Contributions summary:Sarthak primarily focused on updating and improving the `tflite_transfer_converter.py` script, likely related to the conversion of TensorFlow models to TensorFlow Lite format, specifically for transfer learning. They addressed issues related to custom operations and the `SELECT_TF_OPS` target. They also made changes to the `keras_model_head.py` file, supporting the Functional API with InputLayer and refining gradient computations on trainable variables. Furthermore, they made minor code cleanup changes.
tensorflow
sarthak-chakraborty/PedalPals_Android
Apr 2020 - Mar 2021
Contributions:17 commits, 16 pushes, 1 branch in 11 months
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.