Caroline Chen
Founding Engineer at Runhouse
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolCaroline Chen is a founding engineer based in New York with 10 years of experience building production systems and machine learning infrastructure. She helped drive PyTorch Audio development at Meta, contributing backend audio I/O features and implementing a CTC decoder with support for KenLM and custom language models—work that underscores her strengths in speech recognition pipelines and performance-sensitive engineering. At Runhouse she continues to blend systems engineering and ML productization, bringing startup grit to scalable infra challenges. Her background includes internships across quantitative trading, protein AI, and large-scale research at Intel, reflecting a rare mix of applied research, low-latency systems, and ML. She holds a CS degree from MIT and is comfortable moving between C++/Python systems code and ML model integrations—often improving testing and tooling as she ships features.
10 years of coding experience
3 years of employment as a software developer
Bachelor's degree Computer Science, Bachelor's degree Computer Science at Massachusetts Institute of Technology
Spanish, Chinese, English
Github Skills (8)
pytorch10
machine-learning10
speech-recognition10
audio-processing10
python10
testing9
cprogramming-language9
c-language9
Programming languages (3)
Github contributions (5)
pytorch/audio
Jan 2021 - Dec 2022
Data manipulation and transformation for audio signal processing, powered by PyTorch
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:2 releases, 731 reviews, 169 commits in 1 year 10 months
Contributions summary:Caroline primarily contributed to the development and enhancement of the torchaudio library, focusing on audio processing and machine learning components. Their contributions included adding functionality to the "sox_io" backend, specifically the addition of a "dtype" argument to the save function. Furthermore, they implemented and integrated the CTC (Connectionist Temporal Classification) decoder, including supporting both KenLM and custom language models, demonstrating proficiency in implementing speech recognition pipelines and potentially in optimizing the performance of these pipelines. They also updated and improved the testing framework.
pytorchaudio-processingmanipulationpythonsignal
run-house/notebooks
Feb 2024 - Dec 2024
Runhouse example notebooks.
Contributions:8 reviews, 26 PRs, 36 pushes in 10 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.