Nikita Demashov
AI Software Engineer at YADRO
Nizhny Novgorod, Nizhny Novgorod Oblast, Russia
Summary
🤩
Rockstar🎓
Top SchoolNikita Demashov is an AI Software Engineer with six years of hands-on experience building and optimizing inference pipelines, currently driving AI development at YADRO. He has deep practical expertise in low-precision transformations and debugging for production-grade inference, evidenced by contributions to OpenVINO—one of the leading open-source toolkits for AI deployment—where he improved MoveFakeQuantize, multi-channel support, and fixed YOLOv5-related issues. His background includes an internship at Intel on deep learning software and a progression from Junior AI Developer to his current role, showing rapid technical growth and ownership. Trained in Mathematics and Computer Science at Высшая Школа Экономики, he blends strong theoretical foundations with pragmatic engineering. Notably, he focuses on making quantization and graph-transform passes robust for real-world models, a niche that materially reduces model size and latency in deployment. Based in Nizhny Novgorod, he brings both open-source collaboration experience and production-facing ML engineering to edge and enterprise AI projects.
6 years of coding experience
4 years of employment as a software developer
Бакалавр, Mathematics and Computer Science, 8, Бакалавр, Mathematics and Computer Science, 8 at Высшая Школа Экономики
Github Skills (8)
c-language10
deep-learning10
cprogramming-language10
ai10
openvino10
inference9
computer-vision9
python6
Programming languages (6)
Github contributions (5)
openvinotoolkit/openvino
Sep 2021 - Feb 2022
OpenVINO™ is an open source toolkit for optimizing and deploying AI inference
Role in this project:
ML Engineer
ML EngineerContributions:41 reviews, 12 commits, 27 PRs in 5 months
Contributions summary:Nikita primarily worked on low-precision transformation (LPT) within the OpenVINO toolkit, focusing on optimizing and deploying AI inference. Their contributions involve modifying and debugging the MoveFakeQuantize transformation, including fixing constants and supporting multi-channel operations. The user's commits also address bugs related to Concat transformations and fix Yolo v5 issues.
inference-enginepytorchmodel-optimizerdeep-learninggpu
ndemashov/openvino
Jul 2021 - Apr 2022
OpenVINO™ Toolkit repository
Contributions:271 pushes, 28 branches in 8 months
pytorchdeep-learninggpuopenvino-toolkitcomputer-vision
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.