Summary
Jinwoong Kim is a Machine Learning Engineer with 11 years’ experience building high-performance ML systems, from multi-tenant GPU schedulers to production LLM services. He has designed and operated large-scale GPU clusters (H100 SXM5) with NVLink/NVSwitch/InfiniBand, optimized GPU utilization via NVIDIA MIG and MPS, and built Kubeflow-based ML platforms that integrate with complex service meshes and security controls. As a former CTO and co-founder, he led teams to deliver edge and cloud anomaly-detection products and end-to-end ML pipelines, blending product leadership with hands-on systems engineering. His work at NAVER and Toss includes serving GPT‑3 scale models in production and publishing detailed engineering write-ups and conference talks, reflecting both deep research roots (PhD) and pragmatic deployment experience. Based in Seoul, he’s notable for squeezing higher utilization from GPU infrastructure while maintaining secure, observable ML services.
10 years of coding experience
5 years of employment as a software developer
Bachelor’s Degree, Computer Engineering, Bachelor’s Degree, Computer Engineering at Chungbuk National University
Doctor of Philosophy (Ph.D.), Computer Science, Doctor of Philosophy (Ph.D.), Computer Science at Ulsan National Institute of Science and Technology
English, Japanese, Korean