Andrei Andrusenko
Senior Applied Scientist at NVIDIA
Yerevan, Armenia
Summary
🤩
Rockstar🎓
Top SchoolAndrei Andrusenko is a Senior Applied Scientist at NVIDIA and a PhD candidate in Computer Science at ITMO University, with six years of experience building state-of-the-art ASR and DSP systems. He designs and ships end-to-end speech recognition models (CTC, RNN-T, seq2seq) for NeMo while drawing on prior work creating hybrid ASR solutions across many languages at STC Group. His research and engineering blend produced prize-winning results in VOICES, CHiME-6 and GRAM VAANI challenges and peer-reviewed publications in INTERSPEECH and other venues. An active contributor to the ESPnet community, he has improved CHiME6 data pipelines and guided source separation scripts, reflecting a practical focus on robust data preprocessing and MLOps for audio. Based in Yerevan, he pairs academic rigor with production-grade implementation, often surfacing subtle dataset and pipeline fixes that materially improve model performance.
7 years of coding experience
5 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at ITMO UniversityMaster's degree, Computer Systems Networking and Telecommunications, Master's degree, Computer Systems Networking and Telecommunications at Peter the Great St.Petersburg Polytechnic University
Github Skills (14)
sh10
machine-learning10
bash10
speech-recognition10
script10
shell10
python10
scripting10
data-engineering10
docker4
pytorch4
kaldi4
dockers4
chainer3
Programming languages (2)
Github contributions (5)
espnet/espnet
Jun 2020 - Jul 2020
End-to-End Speech Processing Toolkit
Role in this project:
MLOps Engineer
MLOps EngineerContributions:6 commits, 3 comments in 1 month
Contributions summary:Andrei's commits primarily focused on modifying and enhancing scripts related to the CHiME6 dataset within the ESPnet framework. They added and updated scripts for data preparation, particularly for guided source separation (GSS) enhancement techniques. The user made adjustments to the GSS settings, implemented installation procedures for dependencies, and corrected data processing steps. These changes suggest a focus on refining data pipelines and processing for audio analysis.
speech-processingdeep-learningend-to-endchainerpytorch
andrusenkoau/espnet
Mar 2021 - Aug 2022
End-to-End Speech Processing Toolkit
Contributions:16 commits, 16 PRs, 34 pushes in 1 year 4 months
speech-processing
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.