Chinnadhurai Sankar
Research Scientist at Meta
Menlo Park, California, United States
Summary
🤩
Rockstar🎓
Top SchoolChinnadhurai Sankar is a research scientist and leader with a decade of experience building and deploying NLP and generative AI systems, currently at Meta and recently Research Lead at SliceX AI where he developed techniques to produce multiple lightweight LLM slices for cloud and edge inference. He holds a PhD in deep learning from Université de Montréal and a track record of influential research—30+ top-tier publications, SIGDIAL and EACL best paper honors, and program roles at NeurIPS, ICML and ACL. His work spans foundational models, continual learning, conversational AI and efficient transformers, with practical impact such as high-throughput LLM slices demonstrated on TensorRT-LLM. An early contributor across industry labs (Google Brain, Microsoft Research, Twitter, Qualcomm) he pairs production-focused engineering with rigorous research, and even contributes to open-source maintainability work (Theano doc improvements) that improve usability for the community.
10 years of coding experience
5 years of employment as a software developer
Indian Institute of Technology MadrasDoctor of Philosophy (PhD) Deep learning, Doctor of Philosophy (PhD) Deep learning at Université de MontréalMaster of Science Electrical and Computer Engineering, Master of Science Electrical and Computer Engineering at Purdue University
Tamil, English
Github Skills (4)
code-documentation10
documentation10
theano10
python9
Programming languages (2)
Github contributions (5)
Theano/Theano
Mar 2016 - Feb 2017
Theano was a Python library that allows you to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficiently. It is being continued as PyTensor: www.github.com/pymc-devs/pytensor
Role in this project:
Technical Writer
Technical WriterContributions:50 commits, 9 PRs, 13 comments in 11 months
Contributions summary:Chinnadhurai's commits primarily focus on adding and updating documentation strings within the Theano codebase. These changes involve adding docstrings to various files and functions within the `gof` module, specifically concerning the `destroyhandler`, `fg`, `cc`, `vm`, and `op` files. Additionally, the user has documented the `cmodule`, `link`, and `toolbox` files with detailed docstrings. This indicates a strong focus on improving code readability and maintainability through comprehensive documentation.
python-librarymathmulti-dimensionalpythonevaluate
chinnadhurai/ParlAI
Feb 2019 - Aug 2020
A framework for training and evaluating AI models on a variety of openly available dialogue datasets.
Contributions:33 commits, 31 pushes in 1 year 6 months
nlpdeep-learningdatasetmachine-learningtraining
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.