Brian Yan
Graduate Research Assistant
United States
Summary
🤩
Rockstar🎓
Top SchoolBrian Yan is a graduate research assistant and PhD student in Speech and Language Processing at Carnegie Mellon University, specializing in deep learning for speech and NLP. He brings six years of cross-industry experience spanning strategy consulting at Accenture to hands-on ML engineering roles at SoundHound, Tencent, and a research internship at Meta FAIR. His research and open-source work includes contributing a GTN-based CTC loss implementation to the widely used ESPnet speech toolkit, reflecting both practical engineering and novel algorithmic development. With an academic foundation in economics and a computer science minor from the University of Chicago, he combines quantitative rigor with production-minded ML expertise. Brian’s background in strategy and product-facing projects gives him a rare ability to bridge research, engineering, and real-world deployment constraints. He is based in the United States and focused on advancing robust speech models that translate research into embedded and scalable systems.
6 years of coding experience
2 years of employment as a software developer
Doctor of Philosophy - PhD Speech and Language Processing, Doctor of Philosophy - PhD Speech and Language Processing at Carnegie Mellon UniversityBachelor of Arts (B.A.) Economics Major with Computer Science Minor, Bachelor of Arts (B.A.) Economics Major with Computer Science Minor at University of Chicago
English, Chinese, French
Github Skills (5)
pytorch10
machine-learning10
speech-recognition10
python9
deep-learning9
Programming languages (4)
Github contributions (5)
espnet/espnet
Dec 2020 - Dec 2022
End-to-End Speech Processing Toolkit
Role in this project:
ML Engineer
ML EngineerContributions:27 reviews, 64 commits, 20 PRs in 2 years 1 month
Contributions summary:Brian primarily contributed to the implementation of a GTN (Graph-Transformation Network) based CTC (Connectionist Temporal Classification) loss function. Their work involved adding the GTN-based CTC loss to the existing CTC framework, defining the forward and backward passes for the loss calculation using the GTN library. The contributions included integrating the GTN CTC loss function, adding it as an option for loss calculation, and adding documentation and unit tests.
speech-recognitionspeech-separationchainerspoken-language-understandingspeech-processing
brianyan918/brianyan918.github.io
Dec 2020 - Nov 2024
Github Pages template for academic personal websites, forked from mmistakes/minimal-mistakes
Contributions:123 pushes in 3 years 11 months
templategithub-pages-templatemmistakesmistakesjekyll
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.