Summary
Gyubin Son is a data scientist and AI engineer with 10 years of experience specializing in large language models and visual-language systems, currently driving LLM and VLM work at Woowa Bros. His background spans cutting-edge pre-training and fine-tuning at NAVER HyperCLOVA, production ML systems and quantized on‑device inference at MUSINSA, and multimodal research at Samsung Research. He builds end-to-end solutions—data pipelines, custom benchmarks, PEFT trainers, and containerized batch inference orchestration—that move models from experiments to cost‑efficient production. Notably, he has fine-tuned small LLMs to GPT‑4.1–level performance on domain data and deployed QAT+QLoRA quantized models for low-cost inference. A former Marine Corps sergeant with dual business and engineering education, he combines disciplined operational rigor with cross-disciplinary product engineering.
10 years of coding experience
7 years of employment as a software developer
Master's degree Industrial Engineering, Master's degree Industrial Engineering at Korea University
Bachelor of Business Administration (BBA) Business Administration and Management General, Bachelor of Business Administration (BBA) Business Administration and Management General at Yonsei University
Computer Science, Computer Science at 42 Silicon Valley
Korean, English