I am a research scientist at Inria, Soda team and a lecturer at École polytechnique and ENS Paris where I teach competitive programming and deep learning. My research interests are recommender systems, generative models and AI in education (Kandemir et al. 2024), notably using reinforcement learning (Vassoyan et al. 2024).
We are hiring engineers, postdocs and PhD students, as we were awarded a IA-Cluster chair about ATLAS: AI for teaching and learning at scale at Paris-Saclay University & an ANR JCJC grant. You can either wait for a more precise job offer description, or directly reach out with a CV.
Feel free to contact me by mail (PGP key) or X.
Training a student model on student data can be used to optimize teaching
using reinforcement learning.
I am also interested in synthetic tabular data generation (Vie*, Rigaux*, and Minn 2022).
We study applications of collaborative filtering for education (~
multidimensional item response theory), drug repurposing (with Clémence Réda) and diversity of
recommendations (pass Culture, Kyoto U).
Our paper Diversified recommendations of cultural activities with
personalized DPPs [slides] has
been accepted to RecSys workshop RecSoGood 2025.
✅ I worked on the government project Pix. This is free software that certifies
the digital skills of all French citizens
✅ We founded the Girls Can Code!
schools for K-12 girls running since 2014
✅ See my projects and CV
See all publications / My Scholar page