Anton Peganov
Senior Software Engineer at NVIDIA
Munich, Bavaria, Germany
Summary
🤩
Rockstar🎓
Top SchoolAnton Peganov is a Senior Software Engineer based in Munich with nine years of experience building production ML and inference systems at NVIDIA, focused on deployable speech, translation, and LLM services. He has contributed to flagship open-source AI tooling (e.g., NeMo) where he improved perplexity and loss metrics, and has hands-on experience across Python, Docker, CI/CD, and occasional C++ and Go. At NVIDIA he moved from research internships into full-time roles delivering microservices, token-counting endpoints, and inference microservice orchestration, combining deep learning expertise from an MIPT Master’s in Deep Learning with practical engineering rigor. He also taught CS and ML courses earlier in his career, which helps him explain complex systems and drive quality across docs, tests, and refactors.
8 years of coding experience
7 years of employment as a software developer
Master's degree, Deep Learning, Master's degree, Deep Learning at Moscow Institute of Physics and Technology (State University) (MIPT)
Github Skills (9)
neural-network10
pytorch10
machine-learning10
nlp10
deep-learning10
large-language-models10
python10
generative-ai10
metric10
Programming languages (6)
Github contributions (5)
NVIDIA/NeMo
Nov 2020 - Nov 2022
A scalable generative AI framework built for researchers and developers working on Large Language Models, Multimodal, and Speech AI (Automatic Speech Recognition and Text-to-Speech)
Role in this project:
ML Engineer
ML EngineerContributions:63 reviews, 2805 commits, 68 PRs in 2 years
Contributions summary:Anton primarily contributed to the development and improvement of perplexity calculations and metrics within the NeMo framework. Their work involved modifying existing code, adding new metric implementations, and fixing bugs related to the computation of perplexity, a key metric for evaluating language models. They also made updates to documentation, code style and implemented changes in tests and refactored code related to metrics. The user also contributed code related to loss metrics.
asrspeech-recognitionnatural-language-processingttsspeaker-diarization
deeppavlov/learning-to-learn
Feb 2018 - Aug 2019
Contributions:580 commits, 307 pushes, 1 branch in 1 year 6 months
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.