About
I work as a data scientist in the fintech organisation of an e-commerce company, and teach financial AI as an adjunct professor at Ajou University’s graduate school of AI.
This blog collects notes on financial AI, credit scoring, LLM applications, and the problems that turn up once any of it reaches production.
Most posts are written in Korean. The Posts(ENG) section carries English versions of selected pieces.
Books
Financial AI in Practice · Manning (MEAP)
Designing and operating AI applications end to end in a financial setting: credit scoring, fraud detection, investment systems. The book is organised around a four-layer framework for weaving machine learning models, graph analysis, and LLM/RAG together at product quality.
Available to read early through Manning MEAP, with a chapter preview as well.
Practical Knowledge Every Junior AI Engineer Needs · Hanbit Media (2026) · Korean
How do you build a system you can trust on top of a component that behaves probabilistically? Five parts and fourteen chapters on crossing the gap between a demo and production: the paths by which data quietly corrupts a system, five layers of deterministic safety net over probabilistic output, how much autonomy an agent should be given, and what happens to cost and observability after deployment. Weighted toward principles that outlast any particular framework.
Published in Korean. Some of the material also appears here in English → The Model Is Smart, So Why Is the System Stupid?