My research develops mathematical tools and algorithms to address both computational intractability and the practical challenges of solving optimization problems at scale. It is organized around three questions:
How can we construct tractable convex approximations to intractable optimization problems? In particular, I develop and analyze approximations to the cone of nonnegative polynomials based on sum of squares polynomials, and use them to tackle optimization problems over this cone. These approximations allow managers, for example, to build expert knowledge into machine learning models used in real-time inventory contract negotiations or in valuing combinations of products in a shopping basket.
How can we identify tractable instances of otherwise intractable optimization problems? In particular, I study how structural and statistical assumptions that naturally arise in supply chain management and social networks can make such problems tractable. These assumptions make it possible to determine where to hold inventory in a supply chain, which firms to incentivize to adopt a new technology, or when anonymized network data can compromise privacy.
How can we make optimization algorithms run faster on large problems? In particular, I develop methods that scale semidefinite programming, a broad class of convex optimization problems, to large instances, and that accelerate two widely used families of iterative algorithms, fixed-point methods and difference-of-convex procedures. These advances make optimization practical in settings where decisions must be made repeatedly or in real time.
INFORMS Harvey J. Greenberg Research Award — 2026
Awarded annually for outstanding research in computation and operations research with the potential to significantly impact practice.
Honorable mention, INFORMS Computing Society Prize — 2026
Awarded annually for outstanding work at the interface of operations research and computer science.
Patrick and Valentine Firmenich Fellow in Business and Society, INSEAD — 2023–2025
Recognises and supports exceptional young academics working at the intersection of business and society.
Finalist, A. W. Tucker Prize — 2021
Awarded every three years for outstanding doctoral theses in optimization worldwide.
IEEE Information Theory Society Paper Award — 2020
Awarded annually for an outstanding publication in information theory across all publications spanning the preceding four calendar years.
INFORMS Optimization Society Young Researchers’ Prize — 2018
Awarded annually for outstanding papers in optimization by early-career researchers.
School of Engineering and Applied Science Award for Excellence, Princeton University — 2017
Awarded to advanced graduate students for outstanding scholarship and research.
INFORMS Computing Society Student Paper Award — 2016
Awarded annually for the best student paper at the interface of computing and operations research.
Médaille de l’École Centrale — 2013
Awarded annually by the Académie des sciences to the valedictorian of École Centrale Paris.
Gordon Wu Fellowship, Princeton University — 2012–2016
Fellowship awarded to outstanding incoming doctoral students in engineering.
MBA Professors To Look Out For — BusinessBecause, 2020.
A Classical Math Problem Gets Pulled Into the Modern World — Quanta Magazine, 2018. An introduction to polynomial optimization and related work on DSOS and SDSOS methods.