I put data where performance is actually decided: federation-level governance, sensor pipelines, modelling the effort itself.
Research data engineer — INSEP, Paris
Tools and fields: Python, SQL, data governance, Power BI, ETL pipelines, GPS and LPS, machine learning, Scikit-learn, R Shiny, force–velocity profiling, Django, REST APIs, training load, Streamlit, exercise physiology.
What I do four fields, one chain
Trained first as an engineer in computer science and statistics, then in the optimisation of athletic performance. I work at the exact boundary between information-systems architecture and elite sport science, a place where you need to write the pipeline and know what the measurement means.
Mapping athlete data flows, writing the data dictionaries and setting the rules that keep a federation's data estate coherent. Without that work, every coaching staff recomputes its own indicator and nothing can be compared to anything.
Building the API connectors and ETL chains that bring heterogeneous feeds (GPS, LPS, tracking, force plates) into a single reference store, with the quality controls that have to come with them.
Quantifying training load, reading force–velocity mechanical profiles, and estimating a young athlete's potential while accounting for relative, biological and training age. Raw performance conflates those three variables every single time.
Making all of it usable by a coaching staff on a Monday morning: dashboards, visual reporting, and training the people who rely on them. An indicator nobody reads is worth nothing.
Career professional experience
Jan. 2025 → present
INSEP / French Basketball Federation — Paris
Aug. 2026 → present
Self-employed — France
Jan. 2024 → Jan. 2025
INSEP — Paris
Education
2025 → 2026
Université Jean Monnet, Sport Science — Saint-Étienne
2020 → 2025
Polytech Lille, ISIA programme — five-year degree
Mar. 2024 → Feb. 2025
Universidad Nacional del Sur — Bahía Blanca, Argentina
Projects two production apps, on video
Recruitment web application for a professional scouting unit: a pool of several tens of thousands of players, a performance rating per position and per season, field reports partitioned by scout, and trajectory projections.
Monitoring application built for the French Basketball Federation: aggregation of load and wellness measures, reporting back to coaching staff, alerts on deviations.
All public repositories: github.com/Clement0242
Lab machine learning you can try live
Paste a job ad: a small home-made model reads it, compares it with my background and returns a score out of 100, with my strengths and my weaknesses. It doesn't flatter, it computes; an AI can then comment on the result, if you ask.
Talks & papers speaking and publications
5 March 2025
Speaker, session “Tools for detection and recruitment”, alongside Adrien Sedeaud and Quentin De Larochelambert (INSEP / IRMES), Cédric Leduc (Paris Saint-Germain), Elie Rambaud (OL Academy), and the French cycling, triathlon, rowing and football federations.
INSEP, Amphi C — Paris, France
Publications
A permanent role, a one-off engagement, or simply a methodological question about a field dataset: write to me. I always reply.
Based in Paris, France. Mobile across France and available for travel.