Matthijs Douze
Research Scientist at Meta
Paris, Ile-de-France
Summary
🤩
Rockstar🎓
Top SchoolMatthijs Douze is a research scientist with 16 years of experience based in Paris, currently working at Meta/Facebook AI Research where he bridges foundational research and high-performance systems. Trained as an engineer in computer science and mathematics, he brings deep expertise in large-scale similarity search and vector retrieval, contributing core C++ and CUDA optimizations to the widely used FAISS library. His background at Inria and long tenure in research engineering reflect a strong foundation in algorithmic rigor combined with practical performance tuning for GPU-backed workloads. Known for shipping benchmarking and backend features that expose real-world trade-offs, he excels at turning theoretical ideas into production-grade, efficient implementations.
16 years of coding experience
10 years of employment as a software developer
Licentiate degree, Mathematics, Licentiate degree, Mathematics at Lycée Pierre De FermatIngénieur, Ingénieur at ENSEEIHT
Stackoverflow
106reputation
13kreached
3answers
0questions
Github Skills (15)
cuda10
gpu-programming10
c-language10
cprogramming-language10
data-structure9
benchmarking9
benchmark9
algorithm9
data-structures9
algorithms9
vim6
indexing6
python6
emacs6
knn6
Programming languages (7)
Github contributions (5)
facebookresearch/faiss
Feb 2017 - Oct 2022
A library for efficient similarity search and clustering of dense vectors.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:1 release, 146 reviews, 251 commits in 5 years 8 months
Contributions summary:Matthijs's commits focused on adding benchmarking scripts for efficient similarity search and clustering of dense vectors, demonstrating expertise in performance analysis and optimization within the context of the Faiss library. Their work involved the modification of core C++ files. The code changes indicate an understanding of CUDA programming and GPU resource management. These contributions include the development of core back-end features and the modification of library-level code.
clusteringsimilarity-search
mdouze/faiss
Apr 2017 - Oct 2022
A library for efficient similarity search and clustering of dense vectors.
Contributions:52 commits, 293 pushes, 134 branches in 5 years 6 months
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.