Ariel Kwiatkowski
Research Engineer at Meta
Paris, Ile-de-France
Summary
🤩
Rockstar🎓
Top SchoolAriel Kwiatkowski is a research engineer in Paris with a decade of experience at the intersection of reinforcement learning and large language model reasoning, currently advancing LLM-RL methods at Meta. He holds a PhD from École Polytechnique and has led open-source RL tooling as a Farama Foundation maintainer, contributing fixes and tests to flagship projects like OpenAI Gym and PettingZoo. His background in physics and hands-on RL work spans building multi-agent environments, implementing PPO and distributed training, and integrating LLMs into human-in-the-loop systems. Ariel combines rigorous academic research with practical engineering—improving type consistency and test coverage in widely used RL libraries—while pursuing AGI-relevant agent architectures.
9 years of coding experience
6 years of employment as a software developer
Master's degree ICT Innovation specialization: Artificial Intelligence and Robotics, Master's degree ICT Innovation specialization: Artificial Intelligence and Robotics at Aalto UniversityDoctor of Philosophy - PhD Artificial Intelligence, Doctor of Philosophy - PhD Artificial Intelligence at École PolytechniqueMaster's degree Autonomous Systems, Master's degree Autonomous Systems at KTH Royal Institute of TechnologyBachelor's degree Physics (Individual studies), Bachelor's degree Physics (Individual studies) at University of Warsaw
English, Polish, French
Stackoverflow
21reputation
54reached
1answer
0questions
Github Skills (13)
gymnasium10
openai-gym10
multi-agent-reinforcement-learning10
python10
reinforcement-learning10
numpy10
testing10
enviroment9
environ9
api9
dev-environment9
environmental9
pygame7
Programming languages (10)
Github contributions (5)
openai/gym
Aug 2021 - Sep 2022
A toolkit for developing and comparing reinforcement learning algorithms.
Role in this project:
Back-end Developer & Test Automation Engineer
Back-end Developer & Test Automation EngineerContributions:135 reviews, 29 commits, 34 PRs in 1 year
Contributions summary:Ariel primarily focused on improving the type consistency and testing of the observation spaces within the `openai/gym` repository. They modified the data types of various environment observations, making them consistent with the `observation_space`. Moreover, the user added and modified existing test code to accommodate the new data types, ensuring the correctness of observation spaces. This work involved fixing formatting, addressing edge cases, and adapting tests for different environment types.
comparingreinforcement-learning-algorithmsdevelopingdeep-learningreinforcement-learning
Farama-Foundation/PettingZoo
Oct 2021 - Nov 2022
An API standard for multi-agent reinforcement learning environments, with popular reference environments and related utilities
Role in this project:Back-end Developer
Back-end DeveloperContributions:17 reviews, 18 commits, 11 PRs in 1 year 1 month
Contributions summary:Ariel primarily contributed to the `pettingzoo` repository by fixing bugs and improving the `pursuit` environment, a multi-agent reinforcement learning environment. Their commits focused on correcting reward sharing logic, refining capture mechanics, and adding rendering features for the environment. These changes involved modifying the `pursuit_base.py` file and updating the environment's internal state management and rendering capabilities.
agentreinforcement-learningreinforcement-learning-agentgymnasiumdeep-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.