RAG Build Lab: From Documents to a Working AI Assistant is a hands-on, builder-first workshop where you’ll turn real files into a grounded, production-minded Retrieval-Augmented Generation (RAG) app—fast. This session is part of AI Guild and follows the same build discipline used in the AI Residency (scope → architecture → evaluation → iteration).

What you’ll do in this lab

• Convert documents (PDFs/FAQs/notes/web pages) into a searchable knowledge base
• Design a clean RAG pipeline: ingest → chunk → embed → retrieve → generate
• Improve answer quality with better chunking, retrieval tuning, and prompt patterns
• Add “trust” features: citations, refusal rules, and basic guardrails
• Create a lightweight evaluation set to measure quality and reduce hallucinations

Who it’s for

Builders, analysts, engineers, product professionals, and founders who want to move from “RAG demos” to a working assistant you can iterate and ship.

What you’ll leave with

A clear project blueprint + a working v0 direction you can publish as a portfolio demo, plus a repeatable checklist you can reuse for future RAG projects.

Part of Challenge

Bring: a laptop and 2–5 documents you want your assistant to answer from.