Moksh Jain
Graduate Student
Montreal, Quebec, Canada
Summary
👤
Senior🎓
Top SchoolMoksh Jain is a PhD student at Université de Montréal and graduate researcher at Mila under Yoshua Bengio, combining deep academic rigor with nine years of hands-on engineering experience in ML and systems. His work spans multi-agent learning, reinforcement learning, and generative models, and he has a strong record of building user-driven software and production-ready ML components. Moksh has contributed notable open-source implementations—such as a PyTorch Deep CFR for OpenSpiel and an OptimisticAdam optimizer integrated into mlpack—that reflect both research insight and practical optimization skills. He has implemented performance-critical CUDA kernels for edge ML at Microsoft Research and improved FastGRNN for low-resource deployments, highlighting expertise across hardware and algorithmic layers. Based in Montreal, he pairs research collaborations at top labs (DeepMind, Mila) with applied internships and a talent for turning theoretical ideas into tested code. A less obvious strength is his focus on trustworthy, reproducible implementations—tests and documentation feature prominently in his contributions.
9 years of coding experience
2 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Université de MontréalBachelor of Technology, Information Technology, Bachelor of Technology, Information Technology at National Institute of Technology Karnataka
Github Skills (27)
algorithm10
pytorch10
c-language10
python10
optimizers10
coregames10
machine-learning10
optimizer10
rnn-model10
machine-learning-algorithms10
n10
reinforcement-learning10
videogames10
deep-learning10
cplus10
Programming languages (11)
Github contributions (5)
microsoft/EdgeML
Sep 2019 - Jan 2021
This repository provides code for machine learning algorithms for edge devices developed at Microsoft Research India.
Role in this project:
ML Engineer
ML EngineerContributions:1 review, 18 commits, 11 PRs in 1 year 3 months
Contributions summary:Moksh contributed significantly to the `microsoft/edgeml` repository, focusing on the development of machine learning algorithms for edge devices. Their work primarily involved implementing and refining CUDA kernels for the FastGRNN model, including forward and backward passes, and incorporating support for different non-linearities and low-rank matrix approximations. Furthermore, the user added batch_first support to the fastgrnn implementation, added sparsify support, and fixed KWS training by modifying the demo script.
classifieredge-machine-learningsensortensorflowmicrosoft
google-deepmind/open_spiel
May 2020 - Dec 2020
OpenSpiel is a collection of environments and algorithms for research in general reinforcement learning and search/planning in games.
Role in this project:ML Engineer
ML EngineerContributions:11 commits, 1 PR, 11 comments in 6 months
Contributions summary:Moksh implemented a Deep CFR (Counterfactual Regret Minimization) algorithm in PyTorch for the OpenSpiel environment, specifically for the game of Kuhn Poker. This involved creating a PyTorch-based implementation of the Deep CFR algorithm, including neural networks for policy and advantage estimation. The user's work includes setting up the necessary dependencies, building the OpenSpiel library, and integrating the PyTorch implementation within the existing framework.
cppmultiagentgamespythondatamining
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.