LlamaIndex RAG Framework
Build data-centric RAG apps with LlamaIndex. Wires up document ingestion across 300+ connectors, vector and hybrid indices, query engines, and agents with multi-modal support for document Q&A and knowledge retrieval.
This skill builds RAG pipelines on LlamaIndex. It ingests documents through connectors, chooses indexing strategies (vector, summary, knowledge-graph), composes query engines with routing and reranking, adds response synthesis and citations, and wraps retrieval in agents for multi-step document Q&A.
llamaindex rag retrieval vector-search query-engine
When to use
Use when building document Q&A, knowledge-base chatbots, or RAG pipelines centered on your own data, especially with many source connectors or multi-modal content.
Examples
Index a knowledge base
Ingest and query docs
Use LlamaIndex to ingest a folder of PDFs and Notion exports, build a vector index, and expose a query engine that returns answers with source citations
Add a router query engine
Route across multiple indices
Set up a LlamaIndex router that sends factual questions to a vector index and summary questions to a summary index