Shyam Sudhakaran

Research Engineer at Autodesk Research

United States
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts

Summary

🤩
Rockstar
Shyam Sudhakaran is a research engineer with eight years of experience building ML-driven systems and scalable backend services across industry and academia. He blends applied machine learning research—evidenced by contributions to MarioGPT level generation—with production software engineering at companies like AWS and LiveMap, spanning full-stack, distributed systems, and graph neural networks. Currently at Autodesk Research and affiliated with IT University of Copenhagen, he bridges experimental research and practical implementation, often improving usability and tooling in open-source projects. He has a track record of turning novel research ideas into working code and datasets, and brings a pragmatic focus on sampling, data generation, and system robustness that accelerates prototyping to production.
code8 years of coding experience
job5 years of employment as a software developer
github-logo-circle

Github Skills (6)

pytorch10
machine-learning10
gpt10
nlp10
python10
transformers9

Programming languages (5)

JavaRustHTMLJupyter NotebookPython

Github contributions (5)

github-logo-circle
shyamsn97/mario-gpt

Feb 2023 - Feb 2023

[Neurips 2023] Generating Mario Levels with GPT2. Code for the paper "MarioGPT: Open-Ended Text2Level Generation through Large Language Models" https://arxiv.org/abs/2302.05981
Role in this project:
userML Engineer
Contributions:4 releases, 1 review, 29 commits in 8 days
Contributions summary:Shyam contributed significantly to the development of the `mario-gpt` project, focusing on level generation using GPT-2 models. Their work included integrating and modifying the model, implementing code for sampling and generating levels, and improving the codebase's overall structure. They added functionality for level trimming and implemented features for time-lapse generation of levels, indicating a focus on model application and refinement. The user also improved usability by fixing print statements and streamlining the import process.
shyamsn97/controllable-ncas

Feb 2022 - May 2022

Contributions:31 commits, 2 pushes in 3 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.
Request Free Trial