Shivakanth Sujit
Deep RL Researcher at Araya Inc.
Montreal, Quebec, Canada
Summary
👤
Senior🎓
Top SchoolShivakanth Sujit is a Deep RL researcher with eight years of experience building sample-efficient, robust learning systems at the intersection of reinforcement learning and large language models. Currently at Araya, he focuses on real-world RL applications for robotics and control, bringing practical engineering experience from Android development and mobile tooling. His work spans model-based and offline RL, representation learning, and evaluation methodologies, with publications at NeurIPS and TMLR and collaborations with DeepMind and Microsoft Research. Earlier roles at Mila, Microsoft Research NYC, and multiple academic labs produced advances like Variational Sparse Gating for long-horizon planning and PPO-based controller tuning for nonlinear process control. He combines strong experimental rigor with systems instincts—evident from contributions to an Android environmental monitoring app—and is keen on fusing LLMs and autonomous agents to build more general-purpose learners. Based in Montreal, he’s open to collaborations that push RL from benchmarks toward real-world impact.
8 years of coding experience
2 years of employment as a software developer
Research Masters Student, Computer Science, Research Masters Student, Computer Science at Mila - Quebec Artificial Intelligence InstituteBachelor of Technology - BTech, Instrumentation and Control Engineering, 9.3, Bachelor of Technology - BTech, Instrumentation and Control Engineering, 9.3 at National Institute of Technology, TiruchirappalliD.A.V. Group of Schools, Chennai
Github Skills (6)
javas10
mobile-development10
android10
java10
retrofit9
dependency-management9
Programming languages (5)
Github contributions (5)
enviroCar/enviroCar-app
May 2019 - Aug 2019
enviroCar Android Application
Role in this project:
Mobile Developer (Android)
Mobile Developer (Android)Contributions:91 commits, 13 PRs in 2 months
Contributions summary:Shivakanth's commits primarily involve modifications to the enviroCar Android application, specifically related to dependency updates, particularly concerning retrofit libraries. The user updated retrofit dependencies and made related changes in the Remote Module, updating import statements and other necessary methods. Additionally, the user has implemented a new layout and helper classes for handling a ListView of checkboxes in the Report Issue screen, as well as adding features to the report issue screen, including device information and time since crash.
android-applicationandroid
shivakanthsujit/VAE-PyTorch
Oct 2019 - Mar 2021
Variational Autoencoders trained on the SVHN and FashionMNIST data-sets implemented in PyTorch
Contributions:5 commits, 2 PRs, 3 pushes in 1 year 5 months
pytorchdata-setsfashionmnistdeep-learningautoencoders
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.