The blog.
6 posts on ragRSS feed
The 7 Levels of Claude Code and RAG: What Level Are You On?
A full map of Claude Code memory and RAG — from auto-memory to agentic graph RAG. Know your level, the trap at each one, and what to do next.
Karpathy's Obsidian RAG + Claude Code: The No-Database Setup
Andrej Karpathy's Obsidian knowledge base uses no vector database, no embeddings, and no retrieval pipeline. Here's how to set it up with Claude Code for a solo operator or small team.
Claude Code + RAG-Anything: Multimodal RAG for PDFs and Charts
RAG-Anything extends LightRAG with multimodal ingestion for scanned PDFs, charts, tables, and images. Here's how it works and how to install it with Claude Code.
Claude Code + LightRAG: Graph RAG for Large Document Sets
LightRAG is an open-source graph RAG system that plugs into Claude Code. Here's when you need it, how it works, and how to install it in minutes.
Gemini Embedding 2: How to Build Multimodal RAG Right
Google's Gemini Embedding 2 lets you embed videos directly into vector databases — but embedding a video isn't the same as being able to analyze it. Here's the architecture most people are missing.
Claude Code + NotebookLM: Free AI Research Stack
Combine Claude Code with Google's NotebookLM to build a free research pipeline that scrapes YouTube, runs RAG-based analysis, and generates deliverables like infographics and slide decks — without spending tokens on the heavy lifting.
