-Present: Freelance AI expert. Here are some of my collaborations:
- ACC (from October 2024 to March 2025, 4d/week)
- It’s a company aiming to build batteries for electric cars
- Use cases vary from predictive maintenance on tabular data to chatbots and Computer Vision
- I held the position of AI Tech lead
- I supervised a team of 3 data-scientists and 1 ML-Ops
- I improved their agile methodology
- I contributed to the popularization of labeling tools to the business team
- I contributed to the framing, technical solution definition and follow-up of 2 critical use cases (one about image quality control and one for an advanced RAG system)
- Cali Intelligences (since June 2024, ponctual expertise):
- It’s a startup building Computer vision solutions for the retail industry
- I provide them my expertise to help them build new AI models and AI-based products
- I developed for them a workflow to accelerate their labeling process
- AI Forensics (from January 2024 to September 2024, 3-4d/week):
- It’s a european non-profit that provides evidence of platforms and chatbots with opaque behaviors (Youtube, TikTok, Microsoft Copilot …) and makes them accountable
- To do so, web-scraping pipelines are run to collect data. Then, after labeling and analysis, research questions can be answered.
- Thanks to these studies, true impact is reached through the medias and regulation instances at the European commission.
- My main mission: I supervised the european project AALLM (Algorithmic Auditing of LLM-based search engines) funded by NGI Search. In particular, I was leading the data labeling process for the EU elections project. This implies defining the whole labeling workflow of Microsoft Copilot’s answers to elections-related questions, developing two labeling interfaces with an open source package, deploying them and collaborating with a data labeling team.
- I also perform ponctual missions for the European commission as an expert to review european projects
- If you are a small or medium sized french company that is thinking of integrating data and AI in its processes, and wishes to have a diagnosis to reach that goal and that is 50%-funded by the BPI, please reach out to me as I am part of the experts selected to run these data and AI diagnosis !
-Feb 2019 – Jan 2024: Lead data scientist, then, head of the AI Research team at Fieldbox, at Bordeaux, France.
-January 2017 – June 2018: Post-doc at the GREYC at Caen, France. I was working with Julien Rabin and Abderrahim Elmoataz on colored 3D point clouds processing with variational approaches (I defined variants of the Total Variation regularization on grids of arbitrary topology). The goal was to improve the quality of LIDAR scans of artwork for instance for heriatge preservation.
-September – December 2016: Postdoc. at University Federico II, Naples, Italy in collaboration with Luisa Verdoliva and Giovanni Poggi at the GRIP.
– September 2015 – October 2016:
- Temporary teaching and research position (ATER) at Université Paris Descartes.
- PhD defense.
– October 2012 – June 2016: PhD candidate at the LTCI, Télécom Paris, supervized by Florence Tupin (Télécom Paris), Loïc Denis (Hubert Curien Lab. at St-Etienne) and Charles-Alban Deledalle (former CNRS researcher at IMB in Bordeaux and currently research scientist at Brain Corporation in San Diego).
– 2012: Master of Mathematics Vision and Statistical Learning (MVA) at ENS Paris-Saclay (Cachan), France.
– 2010: Bachelor in Applied Mathematics at Université Paris Cité (former Paris Descartes).
For more details here are several versions of my resume: