News

November 2026. New preprint with João P. G. Ramos on the stability of the Prékopa–Leindler inequality: “A height transportation approach to the quantitative Prékopa–Leindler inequality” (arXiv).

As AI tools now make it much easier to build interactive material, we tried a new way of presenting this paper with a companion website Readers can go into as much or as little detail as they like and explore applications.

We think this paper suits this format well because the proof is modular. It is most intuitive when all functions are log-concave, and by incrementing the arguments little by little we can reach the general case.

October 2026. Our advanced course with João P. G. Ramos on “Transport methods for quantitative geometric inequalities”, has been accepted for the 36th Brazilian Colloquium of Mathematics. It will take place at IMPA, 26–30 July 2027. We will soon prepare some notes. Read the course abstract.

Research: understanding the landscape

Understanding an optimization problem starts with understanding its landscape. In Euclidean spaces, derivatives often reveal this geometry. In infinite-dimensional spaces, functional and geometric inequalities can play a similar role. I like to study the rigidity and quantitative stability encoded by these inequalities, and how they help us understand variational problems and their numerical schemes.

From Rio to Paris

Before joining ENSAE/CREST, I was a postdoctoral researcher in the Optimal Transportation team at the Lagrange Mathematical and Computational Research Center, led by G. Carlier, A. Figalli, Q. Mérigot, and F. Santambrogio.

I completed my PhD at Université Paris-Dauphine and the Mokaplan Inria team, supervised by Antonin Chambolle and Vincent Duval. Before that, I studied at EMAp FGV in Rio de Janeiro under the guidance of Maria Soledad Aronna.