Jongย Kim
Member Of Technical Staff at OpenAI
San Francisco, California, United States
Summary
๐คฉ
Rockstar๐
Top SchoolJong Kim is a Member of Technical Staff at OpenAI with 16 years of experience building large-scale, distributed systems and production-grade ML infrastructure. He brings deep expertise in audio and music technologyโhaving built real-time polyphonic synthesis at Spotify and contributed to prominent open-source projects like OpenAI's Whisper and CLIPโwhile also designing recommender and stream-processing platforms at Kakao and Pandora. Jong combines low-latency systems engineering (Scala/Java/C++) with machine learning model work (pitch estimation, speech recognition), enabling end-to-end solutions from model research to scalable deployment. He has a strong academic foundation with a PhD in Music Technology and practical experience optimizing CI/CD, Spark jobs, and fault-tolerant services. Notably, his contributions to Whisper improved decoding and language handling, reflecting an ability to enhance core model behavior as well as system-level performance.
16 years of coding experience
4 years of employment as a software developer
Master of Science, Computer Science and Engineering, 7.76/9.00, Master of Science, Computer Science and Engineering, 7.76/9.00 at University of MichiganDoctor of Philosophy (Ph.D.), Music Technology, 3.76/4.00, Doctor of Philosophy (Ph.D.), Music Technology, 3.76/4.00 at New York UniversityBachelor of Science, Electrical Engineering, 3.83/4.30, Bachelor of Science, Electrical Engineering, 3.83/4.30 at Korea Advanced Institute of Science and Technology
Korean, English, Japanese
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (25)
algorithm10
tokenize10
pytorch10
algorithms10
python10
model-driven10
machine-learning10
implement10
model-building10
keras10
tokenizer10
deep-learning10
tensorflow10
computer-vision10
speech-recognition10
Programming languages (10)
Github contributions (5)
openai/whisper
Sep 2022 - Jan 2023
Robust Speech Recognition via Large-Scale Weak Supervision
Role in this project:
Back-end Developer
Back-end DeveloperContributions:11 reviews, 52 commits, 166 PRs in 4 months
Contributions summary:Jong primarily focused on modifying the `whisper/transcribe.py` file, suggesting a role in the core functionality and logic of the speech recognition system. Their changes involved altering model defaults, renaming variables, and adding features such as the `condition_on_previous_text` flag and improved language handling, reflecting a contribution to the overall performance and configuration options of the transcription process. They also made related adjustments to the decoding and tokenization modules.
recognitionspeech-recognitionscalelarge-scalesupervision
marl/crepe
Apr 2018 - Oct 2022
CREPE: A Convolutional REpresentation for Pitch Estimation -- pre-trained model (ICASSP 2018)
Role in this project:ML Engineer
ML EngineerContributions:1 release, 60 commits, 21 PRs in 4 years 6 months
Contributions summary:Jong contributed to the CREPE project by adding Travis CI configuration files, indicating a focus on continuous integration. They implemented the Viterbi decoder option for smoothing the pitch curve, which enhanced the model's output. Furthermore, they refactored the project into a module and introduced options such as model capacity selection and a step-size parameter, signifying contributions to the project's usability and feature set. The user also addressed bugs related to HMM transition prior and normalization processes.
pytorchrepresentationpre-trained-modelpre-trainedconvolutional
Find and Hire Top DevelopersWeโve 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.