local-ai-packaged

Why we included this project

This is a docker compose template for people who want to stand up a local AI stack without wiring a dozen containers by hand. It pulls together Ollama for local models, Open WebUI as the chat front end, and n8n for agent workflows, with Supabase handling the database and vector store, Flowise for no-code agents, Neo4j for knowledge graphs, SearXNG for private web search, and Langfuse for tracing. Pre-built RAG workflows ship in the repo, so you can import a working agent setup instead of building one from scratch. GPU profiles cover Nvidia and AMD, with CPU-only and Apple Silicon options, so the same template fits whatever hardware you already have. It is not a finished product so much as a documented starting point, and it does expect you to copy .env.example and add a few Supabase variables before everything starts cleanly.