Woonhyuk Baek
Seongnam-si, South Korea
Summary
🤩
Rockstar🎓
Top SchoolWoonhyuk Baek is an experienced AI researcher and engineering leader with 11 years crafting large-scale language and multimodal models, most recently driving KoGPT and Coyo-700M efforts at Kakao Brain. He combines deep research skills (Master's in Computer Vision) with product-grade engineering—designing data curation pipelines, scalable pre-/post-training systems, and a parallel LLM pipelining architecture that powered an AI search service to 2M monthly active users. As co-founder and AI lead at Open Research Inc., he translated long-context LLM research into a production-grade, low-latency search engine, demonstrating a knack for turning LLM theory into real-world impact. An active open-source contributor, he improved KakaoBrain’s KoGPT repo with practical inference optimizations (including float16 support) and clearer user documentation. Based in Seongnam-si, South Korea, he blends hands-on model engineering with cross-functional leadership across research and production teams. Colleagues describe him as an “AI explorer” who seeks pragmatic engineering shortcuts that preserve research rigor.
11 years of coding experience
15 years of employment as a software developer
Master's degree Computer Vision in Computer Science, Master's degree Computer Vision in Computer Science at Gwangju Institute of Science and TechnologyBachelor's degree Media Engineering and Computer Science (double major), Bachelor's degree Media Engineering and Computer Science (double major) at Soongsil University
Github Skills (9)
transformers10
generative-model10
pytorch10
gpt10
nlp10
deep-learning10
python10
k9
huggingface9
Programming languages (3)
Github contributions (5)
kakaobrain/kogpt
Nov 2021 - Mar 2022
KakaoBrain KoGPT (Korean Generative Pre-trained Transformer)
Role in this project:
ML Engineer
ML EngineerContributions:4 reviews, 31 commits, 5 PRs in 3 months
Contributions summary:Woonhyuk focused on developing and improving the KoGPT model, a Korean language generative pre-trained transformer. Their contributions included adding examples and usage documentation for the model, updating the project name and structure, and addressing a specific issue related to the model's implementation. They also made changes to the inference code, including incorporating a float16 version of the model and adding model descriptions, enhancing the user experience and model usability.
nlptransformerspre-trainedgenerative-modelgpt3
wbaek/theconf
Nov 2018 - May 2021
Python Package for Managing Configurations
Contributions:6 releases, 60 commits, 22 PRs in 2 years 6 months
pythonconfigurationconfigconfigurationspython-package
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.