KADE.
en
DONGHYUN (KADE) YOUN
AI · CLOUD · INFRA — ON THE WAY TO APPLIED AI

Evolving with AI.

From on-prem racks to clouds, GPUs and NPUs — an applied engineer who has grown alongside AI since its early days, and keeps growing. Now aimed at cities around the world.

FIG. 0 — EARTH, ORTHOGRAPHIC · STATIC · ● footprints

PILLARS · 7

  1. 01 ON-PREM & CLOUD & AI ENGINEERBuilt and ran on-prem and cloud AI infrastructure shaped to each customer: GPU/NPU inference, MLOps, IAM, monitoring.
  2. 02 K-URBANISM · CITY AS AN ORGANISMBuilding the K-Urbanism platform on my own: a macro-scale project that treats the city as a single organism and makes it possible to talk with it.
  3. 03 NPU EXPERIENCEIntegrated NPU-based vision and generative models with FuriosaAI and SKT, and built monitoring tools that make accelerators approachable to developers.
  4. 04 FROM EARLY MODERN AI TO NOWFrom early vision models to LLM deployment, RAG and quantization, across products and funded R&D.
  5. 05 COMMUNICATIONLed 30–100 soldiers as an engineer officer, then an infra team and a Vietnam dev team; exhibited at trade shows in Tokyo and Seoul; competed in the Google Gemini API Developer Competition.
  6. 06 THE PATHEnvironmental engineering → combat engineer officer → a year of 12-hour days teaching himself software → top trainee at Samsung Multicampus → fifth employee of an AI startup, where he led the infra team as a researcher and built full-stack, from infrastructure to software, while it grew past thirty people and overseas offices.
  7. 07 TEACHING50+ hours of enterprise training: Docker/Kubernetes for a financial group, MLOps/Kubeflow for a bootcamp, Git for an education company.
BLOG15 SHEETS →PROJECTS16 →KADE-6 · ASK MY POSTS