Julien Jerphanion
Scientific Software Engineer at QuantStack
Paris, Ile-de-France
Summary
🤩
Rockstar🎓
Top SchoolJulien Jerphanion is a Scientific Software Engineer based in Paris with nine years of experience at the intersection of computer science, applied mathematics, and scientific open-source engineering. He specializes in making high-quality, long-lived tools—tackling tech debt, performance bottlenecks (CPU/GPU profiling), and packaging challenges across the conda ecosystem. Julien has driven substantial algorithmic speed-ups in scikit-learn (up to 250× in core patterns), contributed performance and numerical fixes to SciPy, and improves reproducible scientific distributions via mamba and conda-forge. His work spans backend systems, build automation, and numerical library maintenance, with nearly all contributions visible publicly on GitHub. Equally comfortable auditing complex academic code and writing hermetic production-ready solutions, he values elegant, optimal designs that scale in practice. Outside core engineering, he mentors contributors and occasionally invests in early-stage biotech and ML projects aligned with his technical and societal interests.
9 years of coding experience
3 years of employment as a software developer
M2 Recherche — Mathématiques, Vision, Apprentissage, Applied Mathematics, M2 Recherche — Mathématiques, Vision, Apprentissage, Applied Mathematics at ENS Paris-SaclayDiplôme d'ingénieur en Génie Informatique, Informatique, Mathématiques Appliquées, Diplôme d'ingénieur en Génie Informatique, Informatique, Mathématiques Appliquées at Université de Technologie de Compiègne (UTC)Visiting Student, Mathematics and Computer Science, Visiting Student, Mathematics and Computer Science at National University of Singapore
English, French, German, Spanish
Github Skills (40)
algorithms10
conda-forge10
scipy10
maintenance10
c-language10
docker10
python10
scripting10
testing10
scikit10
sparse-matrix10
machine-learning10
bash10
cmake10
dockers10
Programming languages (27)
Github contributions (5)
scikit-learn/scikit-learn
Jun 2020 - Jan 2023
scikit-learn: machine learning in Python
Role in this project:
Data Scientist
Data ScientistContributions:1 release, 2275 reviews, 143 commits in 2 years 8 months
Contributions summary:Julien primarily contributed to the scikit-learn library by enhancing and refactoring existing code related to machine learning algorithms. Their work included adding extra requirements, moving and removing specific packages and improving dot product order for logistic regression. They also optimized the creation of KDTree and BallTree, improved support for readonly sparse datasets, and corrected documentation. Furthermore, the user implemented and maintained core algorithms.
data-analysispythonstatisticsdata-sciencelearn-machine-learning
mamba-org/mamba
Nov 2021 - Nov 2021
The Fast Cross-Platform Package Manager
Role in this project:Back-end Developer
Back-end DeveloperContributions:10 releases, 233 reviews, 1 commit in 1 day
Contributions summary:Julien primarily focused on fixing bugs and implementing improvements related to parsing and handling configuration options within the package manager. Their work involved parsing timeout values, testing and formatting configuration files, and adjusting internal data structures for handling remote fetch parameters. The user also contributed to removing outdated code and warnings, demonstrating a focus on code quality and maintenance. Additionally, the user was involved in changes within the Python bindings to the package manager.
cppwindowscondapackage-managerlinux
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.