ylokhande2003/ollama-rag-chatbot-node

A minimal Node.js + Express retrieval-augmented generation (RAG) chatbot that answers questions from local text files using Ollama for embeddings and chat completions. On startup it ingests documents/sample.txt, splits the text into chunks, creates embeddings with an Ollama embedding model, stores them in an in-memory vector store, and uses nearest chunks as context for responses from an Ollama chat model. Who uses it: developers building local or private RAG prototypes, demos, or learning projects that rely on Ollama. Key features: - Local Ollama-based embeddings and chat completion - Chunking and in-memory vector search - Simple POST /chat API that returns answers constrained to document context (responds "Not found in documents." when no match) - Configurable via .env (OLLAMA_MODEL, EMBED_MODEL, PORT) - Includes ingest flow on startup and instructions to update the knowledge base by editing documents/sample.txt and restarting the server

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.