The Persyvact project aims to build tools for analyzing high-dimensional and structured data using cutting-edge mathematical and algorithmic developments. The project involves researchers from three labs (TIMC-IMAG, gipsa-lab, LJK) with a total of 21 people. The project organized several scientific events and is supervising three PhD theses applying machine learning techniques to genomic data analysis problems like imputation, phasing, and clustering. If selected, Persyvact2 would continue this work on structured models and algorithms for high-dimensional complex health data analysis.