Dmitry Yashunin
Research Engineer at Google DeepMind
Menlo Park, California, United States
Summary
🤩
Rockstar🎓
Top SchoolDmitry Yashunin is a research engineer and machine learning leader with a PhD in laser physics and over 15 years of software experience, currently contributing to next-generation Gemini models at Google DeepMind. He combines deep research credentials—30+ publications, 6 patents and a TPAMI paper on HNSW nearest-neighbor search—with hands-on engineering, shipping production ML systems for video understanding, recommender engines, 3D perception and embedded computer vision. As a founding engineer and principal ML lead he has built large-scale distributed training and MLOps pipelines, led a 7-person ML team, and delivered models that measurably improved accuracy and throughput while cutting infrastructure costs. His open-source work on HNSW (widely used and heavily starred) and contributions to testing/build processes highlight a pragmatic focus on reliability and cross-platform builds. Based in Menlo Park, he blends academic rigor with startup speed—often turning novel algorithms into scalable products used by major tech companies.
8 years of coding experience
12 years of employment as a software developer
Master's degree Laser Physics, Master's degree Laser Physics at Advanced School of General and Applied PhysicsDoctor of Philosophy - PhD Laser Physics, Doctor of Philosophy - PhD Laser Physics at The Institute of Applied Physics of the Russian Academy of Sciences
Russian, English
Github Skills (8)
c-language10
continuous-integration10
cprogramming-language10
testing10
build-automation9
python9
githubaction-workflow9
github-ci9
Programming languages (2)
Github contributions (5)
nmslib/hnswlib
Nov 2020 - Jan 2023
Header-only C++/python library for fast approximate nearest neighbors
Role in this project:
Backend Developer & DevOps Engineer
Backend Developer & DevOps EngineerContributions:66 reviews, 16 commits, 40 PRs in 2 years 2 months
Contributions summary:Dmitry primarily focused on improving the project's testing and build processes. Their contributions include refactoring existing code within the testing framework, and updating the build process to accommodate Windows. Further, the user added the capability to run cpp tests in the CI environment. Finally, the user added the ability to replace deleted elements during addition.
python-libraryneighborspythonheadernearest
dyashuni/hnswlib
Nov 2020 - Nov 2023
Header-only C++/python library for fast approximate nearest neighbors
Contributions:2 PRs, 194 pushes, 27 branches in 3 years
python-libraryneighborspythoncppheader
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.