LlamaIndex is a framework that abstracts the entire RAG pipeline into a few lines of code. Where earlier chapters built everything manually (chunking, embedding, retrieval, prompting), LlamaIndex handles it all with SimpleDirectoryReader + VectorStoreIndex + query_engine.
This chapter demonstrates LlamaIndex with three fictional company document sets, showing how the framework automatically routes questions to the correct company’s documents.