Saurav Jha
Postdoctoral Research Fellow at Mila - Quebec Artificial Intelligence Institute
Montreal, Quebec, Canada
Summary
👤
Senior🎓
Top SchoolSaurav Jha is a postdoctoral researcher at MILA and an IVADO fellow focused on robust, reasoning, and responsible AI, bringing a decade of experience bridging research and production. He completed a PhD at UNSW Sydney and has practical ML engineering chops from roles at openstream.ai, Tencent, and Sony where he shipped conversational LLM agents, neuro-symbolic verification POCs, controllable multimodal generation, and personalized diffusion models. His research contributions include the first exemplar-free continual learning framework for Vision Transformers, recognized at a CVPR continual learning workshop. Comfortable moving models to production, he has built AWS-deployed NLP pipelines, REST APIs, and enterprise-grade LLM agents. Saurav’s background spans international academia and industry—Erasmus-funded masters in Europe and internships at Inria—reflecting a taste for cross-cultural, multidisciplinary problem solving. Peers describe him as someone who pairs theoretical rigor with pragmatic system design, often surfacing elegant, reproducible solutions to messy real-world ML problems.
10 years of coding experience
2 years of employment as a software developer
Erasmus Mundus Joint Master's Degree, Computer Science, Erasmus Mundus Joint Master's Degree, Computer Science at Universite de LorraineErasmus Mundus Joint Master's Degree, Computer Science, Erasmus Mundus Joint Master's Degree, Computer Science at University of St AndrewsBachelor’s Degree, Computer Science and Engineering, Bachelor’s Degree, Computer Science and Engineering at Motilal Nehru National Institute Of TechnologyUNSW Sydney
English, Nepali, Hindi, Maithili, French
Github Skills (32)
in-progress10
hindi10
language-processing9
datasets9
activity-recognition9
urdu9
nlp8
keras8
robustness8
natural-language-processing8
benchmarking8
machine-learning8
benchmark8
named-entity-recognition8
machine-translation8
Programming languages (2)
Github contributions (5)
srvCodes/continual-learning-benchmark
Jul 2020 - Nov 2021
Benchmarking continual learning techniques for Human Activity Recognition data. We offer interesting insights on how the performance techniques vary with a domain other than images.
Contributions:53 commits, 2 PRs, 52 pushes in 1 year 4 months
benchmarkingrobustnessinterestingbenchmarkcontinual
srvCodes/Cartoon-Face-Detection-and-Recognition
May 2018 - Mar 2019
Contributions:49 commits, 1 push, 1 comment in 10 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.