Yichen Lu
Research Scientist at Pigeon.ai
San Jose, California, United States
Summary
👤
Senior🎓
Top SchoolYichen Lu is a research scientist specializing in conversational AI, multimodal understanding, and large audio/speech language models, currently building voice agents at Anuttacon. With a BS in Statistics & Computer Science from UIUC and an MS in AI from CMU, he brings eight years of hands-on experience spanning research, production ML, and DevOps for speech toolkits. He contributed build and CI/CD improvements to the well-known ESPnet speech toolkit, showing a knack for making research code reliable and deployable. His work has ranged from discrete speech-unit research and audio-visual fusion at CMU to developing semantic search and evaluation pipelines in industry internships. Based in San Jose, he balances deep model design with pragmatic infrastructure—dataset curation, training platforms, and experiment orchestration—to ship state-of-the-art voice agents. An early founder who scaled large community events, he pairs entrepreneurial drive with rigorous academic training.
8 years of coding experience
7 years of employment as a software developer
Bachelor of Science - BS, Statistics and Computer Science, Bachelor of Science - BS, Statistics and Computer Science at University of Illinois Urbana-ChampaignMaster's degree, Artificial Intelligence, Master's degree, Artificial Intelligence at Carnegie Mellon University
English, Chinese
Github Skills (15)
cd10
script10
bash10
build-pipelines10
ci-cd-pipeline10
build-automation10
cicd10
pipeline10
dockers9
slurm9
docker9
dataprep8
data-preprocessing8
speech-recognition6
pytorch5
Programming languages (3)
Github contributions (5)
espnet/espnet
May 2023 - Sep 2023
End-to-End Speech Processing Toolkit
Role in this project:
DevOps Engineer
DevOps EngineerContributions:1 review, 1 PR, 3 comments in 4 months
Contributions summary:Yichen's commits primarily focus on configuring and modifying the build and deployment process within the repository. This is evidenced by changes to `cmd.sh`, which defines the backend used for running jobs, switching from `slurm` to `local` execution. Further, changes to `data.sh` and `asr2.sh` demonstrate modifications to data preparation and experiment execution steps, crucial elements in the CI/CD pipeline. These modifications highlight the user's role in adapting the project's build and testing infrastructure.
speech-recognitionspeech-separationchainerspoken-language-understandingspeech-processing
yichen14/yichen14.github.io
Oct 2022 - Dec 2022
Contributions:38 commits, 72 pushes, 3 branches in 2 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.