I am a research scientist working at Safran, the research center of Safran Group, a French multinational company that designs and manufactures aerospace and defense equipments. I am working on ML and Reduced-Order Modeling technologies to speed-up industrial computations.
I defended my PhD en 2013 and my Habilitation in 2020.
- Physical Reduced-Order Modeling
- Machine learning for mesh-based data
You can find my articles on my Google Scholar profile, my ResearchGate profile or my Hugging Face profile.
You can find my resume on my LinkedIn profile.
We constructed and published various datasets of solutions to physics problems on Hugging Face and Zenodo, in PLAID format (see Software section below). Interactive benchmark are hosted on Hugging Face, with detailed instructions on how to participate.
- Mordicus, a library implementing a data model for Reduced-Order Modeling
- genericROM, a library implementing non-intrusive Physical Reduced-Order Modeling methods, based on Mordicus' data model
- Muscat, a library implementing I/O routines for fields and meshes formats, pre/post tools, and a finite element engine
- PLAID (Physics Informed AI Datamodel), a library proposing an implementation for a datamodel tailored for AI and ML learning of physics problems
- MMGP (Mesh Morphing Gaussian Processes), a library for learning solutions of physics problems under non-parametrized geometrical variability