Bolun Dai
Machine Learning Engineer at New York University
New York, New York, United States
Summary
🤩
RockstarBolun Dai is a research scientist and PhD candidate in Electrical and Computer Engineering with eight years of experience applying reinforcement learning, robotics, and computer vision to real-world systems. Currently at Fauna Robotics in New York, he combines academic rigor from NYU and CMU with hands-on engineering—designing multi-agent simulation environments, implementing state-of-the-art RL algorithms, and developing deterministic testing for RL libraries. An active open-source contributor, he has modernized core APIs and seeding behavior in prominent RL projects like PettingZoo and MiniWorld to improve reproducibility and rendering support. His background in legged dynamics and compliant joint modeling gives him a unique edge in bridging simulated agents and physical robot control. Practical, detail-oriented, and research-driven, he focuses on making sophisticated algorithms reliable and testable in both simulation and hardware.
8 years of coding experience
Github Skills (13)
openai-gym10
environmental10
dev-environment10
environ10
python10
reinforcement-learning10
enviroment10
testing9
multi-agent-reinforcement-learning9
refactoring9
gymnasium9
api8
numpy8
Programming languages (11)
Github contributions (5)
Farama-Foundation/Miniworld
Oct 2022 - Jan 2023
Simple and easily configurable 3D FPS-game-like environments for reinforcement learning
Role in this project:
Back-end Developer
Back-end DeveloperContributions:3 reviews, 44 commits, 14 PRs in 3 months
Contributions summary:Bolun primarily refactored the `gym_miniworld` environment, particularly focusing on updating the `reset` and `step` functions. They removed the `seed()` function calls. This involved updating the outputs of `reset` and `step` to align with gymnasium API, ensuring proper episode handling, and adding the `render_mode` functionality. Modifications were also made to the rendering process.
3dreinforcement-learning
Farama-Foundation/PettingZoo
Apr 2022 - Oct 2022
A standard API for multi-agent reinforcement learning environments, with popular reference environments and related utilities
Role in this project:Backend Developer & Test Automation Engineer
Backend Developer & Test Automation EngineerContributions:1 review, 37 commits, 9 PRs in 6 months
Contributions summary:Bolun primarily focused on adding seed functionality to the reset methods of multiple environments within the `pettingzoo` library, ensuring deterministic behavior. They modified environment reset functions in several files, including `hanabi.py`, `simple_env.py`, and `base_atari_env.py`, to accept and utilize a seed parameter. Additionally, the user updated test files and utility functions to incorporate seed testing and ensure the environments remain deterministic. This included changes to API tests and test runners.
multi-agent-reinforcement-learningapigymnasiumreinforcement-learningmultiagent-reinforcement-learning
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.