Anatolii Talamanov
Software Engineer at Google
Dublin, Ireland
Summary
🤩
Rockstar🎓
Top SchoolAnatolii Talamanov is a software engineer with 9 years of experience specializing in computer vision, deep learning, and AI compiler optimization, currently working in Dublin. He has driven performance-focused development at Intel—building graph-based CV frameworks, ONNX Runtime and OpenVINO backends, and optimizing inference for Intel hardware—and recently joined Google. Anatolii is an active open-source contributor to high-profile projects like OpenVINO and OpenCV, where he enhanced benchmarking, NPU LLM support, KV-cache management, and added G-API rendering primitives for real-world image formats. He combines low-level kernel and backend work with higher-level demo and tooling improvements (e.g., media decoding via oneVPL and super-resolution demos), showing a rare full-stack approach to ML deployment. Notably, he implemented an inference plugin from scratch during an internship and has experience tuning heterogeneous CPU+GPU pipelines for production use. His background blends practical performance engineering with hands-on contributions that make AI inference faster and more deployable across generations of Intel accelerators.
9 years of coding experience
4 years of employment as a software developer
Master's degree, ML Developer, Master's degree, ML Developer at Школа анализа данныхНижегородский Государственный Университет им. Н.И. Лобачевского (ННГУ)
Github Skills (27)
opencv10
performance-analytics10
c-language10
performance-monitor10
image-processing10
machine-learning10
inference10
n10
openvino10
performance-measurement10
deep-learning10
performance-analysis10
ai10
np10
performance-tuning10
Programming languages (6)
Github contributions (5)
opencv/opencv
May 2019 - Nov 2022
Open Source Computer Vision Library
Role in this project:
Back-end Developer & Computer Vision Engineer
Back-end Developer & Computer Vision EngineerContributions:1112 reviews, 118 commits, 135 PRs in 3 years 6 months
Contributions summary:Anatolii contributed to the `opencv/opencv` repository by implementing new render primitives within the G-API module, specifically for handling different image formats like NV12 and BGR. Their work involved modifying existing code and adding new kernels to support drawing operations on various image types. This included incorporating conversion functionalities and integration of drawing primitives for image processing tasks. Furthermore, the user defined operators and implemented a feature for the `FText` rendering.
pythonvisiondeep-learningc-plus-pluscomputer-vision
openvinotoolkit/openvino
Apr 2021 - Mar 2025
OpenVINO™ is an open source toolkit for optimizing and deploying AI inference
Role in this project:ML Engineer
ML EngineerContributions:97 reviews, 29 PRs, 1 branch in 3 years 11 months
Contributions summary:Anatolii primarily focused on optimizing and extending the performance of OpenVINO, an open-source toolkit for AI inference. Their contributions involved modifying benchmark tools for both throughput and latency, specifically adding the ability to specify the device used (e.g., CPU, NPU). They also updated the documentation on integrating the NPU with OpenVINO's GenAI features, including code examples for both Python and C++. The user's work included updates and extensions to the NPU-specific LLM implementation within OpenVINO, with improvements to KV-cache management.
inference-enginepytorchmodel-optimizerdeep-learninggpu
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.