Adil Zouitine
Founding Research Scientist at UMA
Paris, Ile-de-France
Summary
🤩
Rockstar🎓
Top SchoolAdil Zouitine is a founding research scientist based in Paris with eight years of experience building robust reinforcement learning systems for real-world robotics, currently leading control research for UMA’s robots after a stint at Hugging Face. He completed a PhD-focused research track at ISAE‑SUPAERO/IRT Saint‑Exupéry on robustness to model uncertainty and has a strong track record in multi-agent adversarial RL, sim-to-real GANs, and online learning. An active educator and mentor, he’s taught RL and deep learning at multiple French institutions and built CI-backed coursework for hands-on student projects. He contributes to notable open-source projects—improving testing for torchtyping and adding core online-ML features to river—and holds a Kaggle competition expert rank, highlighting his practical ML chops. Colleagues describe his approach as “making reinforcement learning boring”: pragmatic, reliability-driven, and focused on reproducible, well-tested systems.
8 years of coding experience
6 years of employment as a software developer
MSc. Statistics and Computer Science, MSc. Statistics and Computer Science at Université Paul Sabatier Toulouse IIIDoctor of Philosophy - PhD Intelligence artificielle, Doctor of Philosophy - PhD Intelligence artificielle at ISAE-SUPAERO
Stackoverflow
83reputation
8kreached
2answers
2questions
Github Skills (24)
pytorch10
static-typing10
online-machine-learning10
pytest10
python10
testing10
incremental-learning10
dynamic-typing10
statistics10
machine-learning10
implicit-typing10
optional-typing10
strong-typing10
preprocess9
preprocessing9
Programming languages (5)
Github contributions (5)
online-ml/river
Jan 2019 - Oct 2022
🌊 Online machine learning in Python
Role in this project:
Data Scientist
Data ScientistContributions:15 reviews, 110 commits, 28 PRs in 3 years 9 months
Contributions summary:Adil's primary contributions revolve around implementing and integrating statistical methods and machine learning components within the `river` library. They developed new statistical features such as Min, Max, PeakToPeak, and Sum, and also implemented a MinMaxScaler for data preprocessing. Furthermore, the user added functionalities related to quantile calculations and rolling statistics, directly contributing to the library's core functionality for online machine learning.
online-machine-learningpythonincremental-learningmachine-learningonline-learning
patrick-kidger/torchtyping
Apr 2021 - Apr 2021
Type annotations and dynamic checking for a tensor's shape, dtype, names, etc.
Role in this project:QA Engineer / Test Automation Engineer
QA Engineer / Test Automation EngineerContributions:2 reviews, 6 commits, 2 PRs in 2 days
Contributions summary:Adil primarily contributed to the testing infrastructure of the `torchtyping` library. They implemented and expanded upon existing test cases, focusing on the shape and type checking capabilities provided by the library. Specifically, they introduced tests to support the `Any` type for shape dimensions, updated existing test functions, and fixed test naming inconsistencies. This effort directly improved the test coverage and reliability of the library.
dtypetensortype-annotationstensorsnamed-tensors
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.