Edgar Duéñez-guzmán
London, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolEdgar Duéñez-guzmán is an AI-focused scientist and engineer based in London with over a decade of experience in applied and fundamental research and five years of hands-on industry experience. He builds tools to augment human autonomy, bridging research and production by contributing to complex multi-agent reinforcement learning projects like DeepMind's widely-used Melting Pot. Edgar excels at backend development, debugging, and environment tooling—skills demonstrated by fixes to simulation logic, environment builders, and logging in large open-source codebases. His background combines deep technical problem solving with practical engineering, and he often surfaces subtle architecture issues (such as instance-check ordering and substrate mappings) that improve system reliability.
5 years of coding experience
University of Tennessee, Knoxville
English, Spanish
Github Skills (4)
lua10
python10
debugging9
tensorflow5
Programming languages (3)
Github contributions (5)
google-deepmind/meltingpot
Jul 2021 - Jan 2023
A suite of test scenarios for multi-agent reinforcement learning.
Role in this project:
Backend Developer
Backend DeveloperContributions:1 release, 50 reviews, 35 commits in 1 year 6 months
Contributions summary:Edgar primarily contributed to the `meltingpot` repository by addressing code documentation, fixing the order of instance checking to convert DM Specs to Gym Spaces, and decreasing the logging level in base_simulation. Their work involved modifying Lua scripts for game object and simulation logic and Python scripts related to environment building and substrate testing. The user also added a mapping of substrate names to scenarios and fixed logging strings, demonstrating an understanding of the project's architecture and debugging skills.
multi-agent-reinforcement-learningmultiagent-reinforcement-learning
google-deepmind/educational
Apr 2021 - Jul 2021
Contributions:2 commits, 2 PRs, 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.