RAG Architecture

Use when designing or fixing a retrieval system for an LLM application — deciding between RAG, cache-augmented generation and plain long context, chunking strategy, hybrid dense+lexical retrieval, reranking, embedding-model and vector-store selection, metadata filtering and per-tenant access control, incremental ingestion and re-embedding, and the evaluation metrics that tell you whether retrieval or generation is the thing that is broken.

joogy06 Updated 5 repo stars

File contents

joogy06/agent-foundry commit 2b7756413f

Frequently asked questions

npx get-skill add joogy06/rag-architecture