Gergely Szilvasy
Research Engineer at Meta
Greater Lille Metropolitan Area France
Summary
🤩
Rockstar🎓
Top SchoolGergely Szilvasy is a research engineer with 12 years of experience building and scaling safety, integrity, and foundational AI systems at Meta, where he progressed from engineering roles to senior staff and research engineering positions. His background spans large-enterprise engineering leadership at Oracle and long-term software development roles dating back to the late 1990s, giving him a rare blend of product-focused management and deep hands-on systems work. He contributes to high-profile open-source projects such as FAISS and Haxl, improving core build, testing, and vector-search functionality that supports efficient similarity search and robust Haskell tooling. Based in the Greater Lille area, he combines practical expertise in abuse detection and media matching with low-level performance optimizations and cross-platform build fixes—skills that surface in both production safety platforms and research prototypes.
12 years of coding experience
26 years of employment as a software developer
Bachelor of Science (BS), Computer Science, Bachelor of Science (BS), Computer Science at Eötvös Loránd University
Github Skills (11)
cabal10
c-language10
cprogramming-language10
haskell10
testing10
parallel-computing9
data-structure9
build-process9
data-structures9
dependency-management9
openmp8
Programming languages (4)
Github contributions (5)
facebook/Haxl
May 2016 - Sep 2016
A Haskell library that simplifies access to remote data, such as databases or web-based services.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:10 commits, 3 PRs, 2 pushes in 4 months
Contributions summary:Gergely primarily contributed to improving the build process, fixing compilation errors, and enhancing the test suite of the Haxl library. They added missing dependencies to the cabal file, resolved test runner conflicts, and unbroke the cabal test by addressing failing tests and test configurations. Furthermore, the user fixed compilation errors related to missing imports and versioned package dependencies. They also addressed build issues and updated the package version for Hackage.
databasedatabaseshaskell
facebookresearch/faiss
Oct 2022 - Jan 2023
A library for efficient similarity search and clustering of dense vectors.
Role in this project:Back-end Developer
Back-end DeveloperContributions:6 reviews, 5 commits, 86 PRs in 2 months
Contributions summary:Gergely contributed to the FAISS library by implementing and testing index cloning functionality for various index types, including `IndexRefine`, `IndexIVF`, and `IndexHNSW`. They added support for iterable inverted lists, which enable the use of key-value stores within the index. Further contributions include the implementation of range search functionality and optimization through tiling, as well as addressing build issues related to OpenMP on macOS and adding support for ARM64 builds.
clusteringsimilarity-search
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.