Track 35Pro Track~110 min total
AI Second Brain: LightRAG
Deploy a graph-powered AI second brain on your own server. LightRAG auto-extracts entities and relationships from your notes, docs, and session logs. Query your entire knowledge base from Claude Code, Telegram, or any API client. The production-grade second brain for builders who want agents that remember everything.
Prerequisite:
§ The modules
- 01What Is Graph RAG and Why It MattersHow LightRAG combines vector search with knowledge graphs. Why flat files and keyword search fail for growing knowledge bases. The architecture in plain terms.Account
- 02Planning Your LightRAG DeploymentWhat you need to run LightRAG: VPS, local machine, or cloud. LLM and embedding options, storage backends, and architecture decisions before you build.Account
- 03Deploy LightRAG on Your VPSDocker Compose setup, LLM and embedding configuration, health checks, WebUI access, and SSL. From zero to running. Also covers local deployment for those who prefer no cloud.Pro
- 04Feed Your Knowledge Base and Build the GraphIngestion strategies: what to feed, what to exclude, how to structure documents. Watch the knowledge graph build in real time.Pro
- 05Query From Claude Code via MCPInstall the MCP server, configure Claude Code, and start querying your knowledge base with natural language from your terminal.Pro
- 06Access From Telegram and Messaging AppsBuild a custom skill that queries LightRAG from Telegram. Get answers about your projects from your phone.Pro
- 07Automated Ingestion and Session LoggingAuto-push new content to LightRAG. Build skills that ingest automatically. Keep the graph growing without manual effort.Pro
- 08Advanced: Query Modes, Reranking, and MultimodalLocal vs global vs hybrid queries. Reranker setup. Multimodal embeddings for images, PDFs, and audio.Pro
Start module 01 →Finish all six for the Pro Track