Vahid Noroozi
Research Scientist at NVIDIA
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolVahid Noroozi is a research scientist and Ph.D. graduate in Computer Science with nine years of experience building deep learning systems for representation learning, semi-supervised and multi-view learning, recommender systems, and graph embedding. Based in San Francisco, he has been advancing large language modeling, speech and NLP at NVIDIA since 2019 and contributed notable backend and model work to high-profile open-source toolkits like NVIDIA NeMo and OpenSeq2Seq. His contributions include implementing ConvS2S components and enhancing intent/slot pipelines to handle new datasets (e.g., NVIDIA-CAR), demonstrating a blend of research rigor and production-oriented engineering. Trained under Prof. Philip S. Yu at UIC, he combines academic depth with practical ML engineering to take models from novel ideas to scalable tooling.
9 years of coding experience
5 years of employment as a software developer
Bachelor of Science (B.Sc.), Computer Science and Engineering, Bachelor of Science (B.Sc.), Computer Science and Engineering at Shiraz UniversityMaster of Science (M.Sc.), Artificial Intelligence, Master of Science (M.Sc.), Artificial Intelligence at Amirkabir University of TechnologyDoctor of Philosophy (Ph.D.), Computer Science, Doctor of Philosophy (Ph.D.), Computer Science at University of Illinois at Chicago
Persian, English
Github Skills (10)
data-preprocessing10
sequence-to-sequence10
machine-learning10
speech-recognition10
nlp10
deep-learning10
tensorflow10
python10
natural-language-processing10
ml10
Programming languages (2)
Github contributions (5)
NVIDIA/NeMo
Nov 2019 - Dec 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:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:821 reviews, 371 commits, 269 PRs in 3 years 1 month
Contributions summary:Vahid updated and improved several examples relating to intent detection, slot tagging, and domain classification. These updates included code modifications to process and incorporate new datasets, such as the NVIDIA-CAR dataset. The user's work involved modifying existing code files and implementing new functions within the project's collections of natural language processing tools, demonstrating their work in applying and refining these components.
asrspeech-recognitionnatural-language-processingttsspeaker-diarization
NVIDIA/OpenSeq2Seq
Jun 2018 - Aug 2018
Toolkit for efficient experimentation with Speech Recognition, Text2Speech and NLP
Role in this project:Back-end Developer
Back-end DeveloperContributions:156 commits, 14 PRs, 4 pushes in 1 month
Contributions summary:Vahid implemented a Conv seq2seq implementation within the open_seq2seq toolkit. The code changes involve modifications to the ConvS2SDecoder, ConvS2SEncoder, and related configurations, which suggests the development of core components for a sequence-to-sequence model. The user also added position embedding and updated the text layer within the code. These changes indicate a focus on expanding the toolkit's capabilities for speech recognition and NLP tasks.
nlpexperimentationseq2seqspeech-to-textrecognition
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.