What is QMD? A Practical Demo of Local-First AI Search

Описание: In this session, we explore QMD hands-on and try to understand what it does well, where it fits, and why it is useful for local-first retrieval workflows.

We start with a folder of markdown/plain text documents, index it with QMD, and compare basic keyword search with QMD’s retrieval pipeline. Along the way, we look at how QMD combines BM25 keyword search, vector semantic search, query expansion, hybrid retrieval, reranking, and MCP support for agent workflows.

Topics covered:

What QMD is
How to index local documents
CLI-based search and vector search
Using QMD with MCP
BM25 vs semantic search
Hybrid retrieval and reranking
Why grounding agent search in real documents matters

This is a practical explainer for anyone interested in RAG, retrieval systems, local-first AI tools, or better search over markdown/document repositories.