Jocelyn Huang
Research Scientist at NVIDIA
San Jose, California, United States
Summary
🤩
Rockstar🎓
Top SchoolJocelyn Huang is a research scientist at NVIDIA with 11 years of software and machine learning experience, specializing in speech synthesis, G2P, and speech recognition on the NeMo team. A Carnegie Mellon alum with a BS in Computer Science and an MS in Machine Learning, she blends rigorous academic training with production-focused research engineering. Her contributions to the widely used NVIDIA/NeMo framework include extending data layer flexibility for custom function-based generation, accompanied by documentation and tests—evidence of both deep technical skill and attention to usability. Based in San Jose, she pairs experience across internships at Google and Microsoft with robotics and RL work from an ISAAC team stint, giving her a rare cross-domain perspective on speech, language grounding, and scalable AI tooling.
11 years of coding experience
1 year of employment as a software developer
Thomas Jefferson High School for Science and TechnologyComputer Science, Computer Science at Carnegie Mellon University
English, Chinese, Japanese
Stackoverflow
101reputation
0reached
0answers
0questions
Github Skills (6)
neural-network10
pytorch10
python10
documentation9
deep-learning9
generative-ai8
Programming languages (3)
Github contributions (5)
NVIDIA/NeMo
Oct 2019 - Jan 2023
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
Back-end DeveloperContributions:611 reviews, 181 commits, 164 PRs in 3 years 3 months
Contributions summary:Jocelyn primarily modified the `RealFunctionDataLayer` class, introducing the ability to pass in a function argument and defaulting to `torch.sin()`. These changes allow users to define custom functions for data generation within the neural module, demonstrating flexibility in data input. The user also updated documentation, examples, and tests to reflect and support the new functionality of the `RealFunctionDataLayer`, ensuring usability and maintainability. The code adjustments and documentation updates indicate a focus on extending the library and ensuring proper use and functionality.
asrspeech-recognitionnatural-language-processingttsspeaker-diarization
redoctopus/cmu-adv
Nov 2015 - Oct 2017
Contributions:30 PRs, 61 pushes, 9 branches in 1 year 10 months
dungeongamemulti-user-dungeonthinkcmu
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.