A RAG chatbot is a starter project where a chatbot answers questions from your own documents by first searching them, then giving the best passages to a language model.
In plain words
The project has two stages. First, indexing: load your documents, split them into smaller chunks, turn each chunk into an embedding, and store the embeddings in a vector store. Second, at question time: search the store for chunks similar to the question and pass them to a model, which writes the answer from them. The idea comes from retrieval-augmented generation, which pairs a model with a searchable index of text. Official tutorials from LangChain and LlamaIndex walk through building one.
A quick example
LlamaIndex's starter tutorial downloads a Paul Graham essay into a data folder, builds a vector index over it, and gives an agent a tool that searches the essay. The agent can then answer questions about the essay, and can also switch to a calculator tool for maths.
Where you'll meet it
- Question answering over documentation, as in LangChain's tutorial.
- Agents that mix document search with other tools.
- Semantic search over knowledge bases.
Related
Sources
- DocsRetrieval Augmented Generation (RAG), LangChain documentation · read 29 Sept 2026
- DocsStarter Tutorial, LlamaIndex documentation · read 29 Sept 2026
- PaperRetrieval-Augmented Generation for Knowledge-Intensive NLP Tasks, Lewis et al. (arXiv) · read 29 Sept 2026
- DocsRetrieval, OpenAI API documentation · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.