Andrei Nesterov
Head Of The Software Development Group at Foxford
Moscow, Moscow City, Russia
Summary
👤
Senior🎓
Top SchoolAndrei Nesterov is a seasoned software leader with 15 years of experience building scalable, real-time education platforms and cloud-native infrastructure from the ground up. As Head of the Software Development Group at Foxford he led teams that delivered ultra-low-latency WebRTC streaming, event-driven collaboration features, and highly profitable classroom products while cutting storage and bandwidth costs through architecture and tooling changes. He introduced GitOps and end-to-end ML engineering into the company, including a real-time toxic comment classifier, and has hands-on ML contributions to notable open-source projects like Google’s Trax. Equally comfortable in backend systems—evidenced by HTTP protocol work in cowlib—and in product-driven engineering, he combines tight technical ownership with measurable business impact, such as a 7.8% net profit increase from a single feature set. Based in Moscow, he pairs dual degrees in engineering and economics with a pragmatic focus on operational efficiency and scalable collaboration.
15 years of coding experience
11 years of employment as a software developer
BS, Economics and Management, BS, Economics and Management at Voronezh State University of Engineering Technology (VSUET)BS, Computer Science and Engineering, BS, Computer Science and Engineering at Voronezh State Technical University (VSTU)
English, Russian
Github Skills (21)
parserator10
parser10
erlang10
python10
http10
machine-learning10
numpy10
deeplearning-ai10
deep-learning10
tensorflow10
jax10
reinforcement-learning9
transformer9
data-structures8
client-websocket8
Programming languages (16)
Github contributions (5)
google/trax
Aug 2020 - Jun 2021
Trax — Deep Learning with Clear Code and Speed
Role in this project:
ML Engineer
ML EngineerContributions:8 commits, 8 PRs, 9 comments in 10 months
Contributions summary:Andrei made several contributions focused on enhancing the Trax deep learning library. They implemented new metrics, including binary cross-entropy loss, macro-averaged F-score, and weighted F-score, expanding the library's evaluation capabilities. Additionally, the user addressed issues related to model checkpoint loading, specifically for SelfAttention layers, and added a new data combinator. The contributions indicate a focus on model development, evaluation, and library functionality.
pytorchspeeddeep-learningnumpyreinforcement-learning
ninenines/cowlib
Feb 2016 - Oct 2016
Support library for manipulating Web protocols.
Role in this project:Back-end Developer
Back-end DeveloperContributions:12 commits, 18 PRs, 13 comments in 7 months
Contributions summary:Andrei contributed primarily to the `cowlib` library, focusing on HTTP-related functionalities. Their work included adding parsing and building methods for HTTP headers such as `Origin`, `Access-Control-Request-Headers`, `Access-Control-Allow-Origin`, `Access-Control-Request-Method`, `Access-Control-Allow-Headers`, `Access-Control-Allow-Methods`, `Access-Control-Allow-Credentials`, `Access-Control-Max-Age`, and `Access-Control-Expose-Headers`. Additionally, they added support for RFC-compliant date formatting. The user's contributions also included writing test cases using Erlang's testing framework (likely `eunit`).
erlangcrystal-langdubhttp2support-library
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.