LLM 실전 워크샵 — 프로덕션 RAG

RAG works great in a demo, but the moment you ship it to a real service something always breaks. The chunking feels off, the embeddings pull up the wrong things, or you skipped evaluation and have no idea when it started getting worse. This workshop is not a lecture — it is a session where we build with our hands together. Open your laptop and we will stand up a small RAG pipeline from start to finish: comparing chunking strategies, how to choose embeddings, and the eval automation that everyone skips the most. Seats are capped at 20. The goal is to work things out on the spot — with the person next to you, and with the host whenever you get stuck. A laptop and a little Python experience is all you need.