Yury Malkov
Researcher at Google DeepMind
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolYury Malkov is a researcher and engineer with a PhD in laser physics and over a decade of experience building high-performance search and computer vision systems. He is the author of HNSW, a widely used approximate nearest neighbor algorithm and core contributor to hnswlib and NMSLIB, work that has been adopted and reimplemented by major companies like Facebook, Microsoft, and Amazon. His career spans research and production roles at DeepMind, OpenAI, Twitter, Samsung AI, and startups, combining theoretical rigor with hands-on engineering for large-scale ML and recommender systems. Yury’s background in experimental physics informs a pragmatic approach to complex systems—optimizing algorithms, serialization, and multithreading for real-world performance. He frequently bridges research and product needs, delivering state-of-the-art solutions from multi-camera 3D pose estimation to transformer-based content understanding.
10 years of coding experience
17 years of employment as a software developer
Doctor of Philosophy (PhD) Laser physics, Doctor of Philosophy (PhD) Laser physics at Institute of Applied Physics of the Russian Academy of Sciences (IAP RAS)Master's degree physics, Master's degree physics at State University of Nizhni Novgorod named after N.I. Lobachevsky (UNN)
Русский, English
Github Skills (25)
algorithm10
code-optimization10
algorithms10
benchmark10
c-language10
approximate10
python10
multithreading10
nearest-neighbors10
benchmarking10
data-structure10
optimisation10
nearest-neighbor-search10
data-structures10
numerical-optimization10
Programming languages (4)
Github contributions (5)
nmslib/hnswlib
Jul 2017 - Jan 2023
Header-only C++/python library for fast approximate nearest neighbors
Role in this project:
Back-end Developer
Back-end DeveloperContributions:9 releases, 82 reviews, 199 commits in 5 years 7 months
Contributions summary:Yury primarily contributed to the development of the hnswlib library, focusing on enhancements and bug fixes related to its core functionality. Their work included adding Python bindings to the C++ code, refactoring existing code, and resolving bugs. Additionally, the user addressed issues with tests, and serialization processes, and also implemented features such as inner product and cosine distance metrics.
python-libraryneighborspythonheadernearest
erikbern/ann-benchmarks
Apr 2016 - May 2020
Benchmarks of approximate nearest neighbor libraries in Python
Role in this project:ML Engineer
ML EngineerContributions:22 commits, 4 PRs, 59 comments in 4 years 1 month
Contributions summary:Yury contributed to the development and benchmarking of approximate nearest neighbor (ANN) libraries in Python. Their work included implementing dummy algorithms for testing and setting up multithreading capabilities for the FALCONN library. They also added functionality to manage CPU affinity and made changes to the NMSLIB library. Furthermore, they were involved in merging branches and updating the main runner script to improve the evaluation process.
pythonnearest-neighbornearestdockerbenchmark
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.