CV

Research Engineer Expert

EDF R&D · OSIRIS, Palaiseau, France

• Led the ground-up redesign and production deployment of Apogène, EDF’s short-term unit commitment software (MILP, branch-and-bound, ADMM, RCSPP, Min Cost Flow). 4 versions deployed in production (V2 2018, V3 2020, V4 2021, V5 2025), saving several tens of millions of euros per major release.
• Managed a team of 10–20 engineers and researchers across the full lifecycle: mathematical modeling, algorithm design, C++ implementation, CI/CD (GitLab), HPC deployment (Slurm, Singularity).
• Integrated ML-based branching heuristics (RL, GNN) into the solver loop; results published at NeurIPS 2025 and AAAI 2026.
• Designed and built the optimization platform for DREEV (EDF × Nuvve joint venture): EV fleet dispatch across NEBEF, Spot, Intraday, and FCR markets (MILP, Lagrangian decomposition). DREEV became first French EV aggregator accredited for FCR by RTE (2022). Ongoing research on Frank-Wolfe decomposition methods as a replacement for Lagrangian decomposition in legacy algorithms.
• Earlier work (2008–2012): task scheduling for maintenance planning, LNG terminal planning & scheduling, combustion and emission optimization for coal power plants.

Postdoctoral Fellow

École Centrale de Paris, France

• Combinatorial algorithms for medical imaging, with Nikos Paragios.

Postdoctoral Fellow

University of Western Ontario, Canada

• Graph cuts and combinatorial algorithms for computer vision, with Yuri Boykov.

PhD in Computer Science and Mathematics

École Nationale des Ponts et Chaussées (ENPC), France

Computer Vision

• Distinction. Dissertation: On Some Extensions of Level Sets and Graph Cuts & Their Applications to Image and Video Segmentation

Master MVA — Mathématiques, Vision et Apprentissage

ENS Paris-Saclay (formerly ENS de Cachan), France

Computer Vision

IndusRO'2022 Prize

ROADEF — French Operational Research Society

Awarded for the industrial impact of Apogène on EDF’s power plant operations.