Franck Charras
Machine Learning Software Engineer at :probabl.
Massy, Île-de-France, France
Summary
🤩
Rockstar🎓
Top SchoolFranck Charras is a Machine Learning Software Engineer with eight years' experience bridging applied mathematics, ML research, and production software. He holds master's degrees in Machine Learning and Applied Mathematics and has led technical, scientific, and people development as co-founder and CTO of a medtech startup. More recently he ran exploratory development and benchmarked GPU and OneAPI compute backends for scikit-learn in a joint INRIA–Intel project, and contributes to scikit-learn’s docs and tests. At present he develops ML software at :probabl., combining hands-on engineering with performance-focused tooling for Python ML ecosystems. Notably, he pairs deep domain knowledge in hospital data and explainability techniques with practical ML engineering to move research prototypes toward robust, accelerated deployments.
8 years of coding experience
7 years of employment as a software developer
Master's degree Applied Mathematics, Master's degree Applied Mathematics at ENS Paris-SaclayMP* Mathematics and Physics, MP* Mathematics and Physics at Lycée Saint LouisMaster's degree Machine Learning, Master's degree Machine Learning at École nationale des ponts et chaussées
French, English
Github Skills (9)
scikit-learn10
machine-learning10
python10
data-science10
documentation10
scikit10
numpydoc9
data-analysis8
testing8
Programming languages (14)
Github contributions (5)
scikit-learn/scikit-learn
Aug 2022 - Nov 2022
scikit-learn: machine learning in Python
Role in this project:
Data Scientist
Data ScientistContributions:66 reviews, 6 commits, 19 PRs in 2 months
Contributions summary:Franck primarily focused on improving the documentation of the scikit-learn library, specifically by ensuring that various dataset loading functions passed numpydoc validation. They addressed docstring issues in several functions, including `fetch_species_distributions`, `fetch_lfw_people`, and functions related to loading SVMlight files. Additionally, the user made improvements to the tests, fixing a uniform random number generator in a KMeans test.
data-analysispythonstatisticsdata-sciencelearn-machine-learning
soda-inria/sklearn-engine-benchmarks
Sep 2023 - Feb 2024
Benchopt suite to compare alternative scikit-learn engines for the same estimators
Contributions:14 reviews, 29 PRs, 142 pushes in 5 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.