Agnès Leroy
GPU Director at Zama
Paris, Ile-de-France
Summary
🤩
Rockstar🎓
Top SchoolAgnès Leroy is a GPU director and engineering leader with a decade of experience translating numerical science into high-performance production software. Trained in mechanical and civil engineering and holding a PhD in computational fluid dynamics, she pioneered GPU-accelerated Smoothed Particle Hydrodynamics during her doctoral work, publishing extensively and winning multiple awards. After contributing to open-source fluid solvers at EDF and building power-grid simulation tools at RTE, she now leads GPU acceleration at Zama to speed up Fully Homomorphic Encryption workloads. She combines deep numerics and C++ expertise with practical systems engineering, moving algorithms from research prototypes to robust, deployable implementations. Based in Paris, she is as comfortable profiling kernels and optimizing memory patterns as she is aligning multi-disciplinary teams around measurable performance goals. An understated strength is her track record of turning complex physics models into maintainable GPU codebases that scale across real-world industrial problems.
9 years of coding experience
10 years of employment as a software developer
Graduate Civil Engineering, Graduate Civil Engineering at Universidade Federal de Minas GeraisGraduate Mechanical Engineering, Graduate Mechanical Engineering at Ecole nationale des Ponts et ChausséesDoctor of Philosophy (PhD) Computational fluid dynamics, Doctor of Philosophy (PhD) Computational fluid dynamics at Université Paris-Est, Ecole Doctorale SIEDiplôme d'ingénieur Génie Mécanique et Matériaux, Diplôme d'ingénieur Génie Mécanique et Matériaux at École nationale des ponts et chaussées
English, French, Portuguese, Spanish
Github Skills (55)
paillier10
crypto10
modular10
cim10
python10
cryptanalysis10
modeling-tool10
ecosystem10
groovy10
llvm10
low-level10
fhe10
power-systems10
java10
boolean10
Programming languages (5)
Github contributions (5)
zama-ai/concrete-core
Oct 2021 - Jan 2023
Concrete Core contains a set of low-level primitives which can be used to implement Fully Homomorphically Encrypted (FHE) programs.
Contributions:9 releases, 546 reviews, 154 commits in 1 year 3 months
cryptographyconcreteprivacy-enhancing-technologiespublic-key-cryptographyrust
agnesLeroy/basic-repo
Mar 2023 - May 2024
Basic Python project to layout
Contributions:6 PRs, 3 comments in 1 year 2 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.