llamaindex

Use this framework to ingest documents from over 300 data sources, structure them into index types like vector or tree, and create a query engine to ask questions over your private data. This enables natural language retrieval for your RAG application.

Connect an LLM to external data using data connectors that ingest files, web pages, databases, and APIs. This framework structures the ingested data into indexes for efficient knowledge retrieval and question-answering.

Yes, you can use OpenAI and Anthropic models. The framework lists both as dependencies to build LLM applications and agents that leverage these models for querying your indexed data.

Ingest internal documentation by connecting to your data sources, then index the loaded documents using vector, list, or tree structures. This creates a query engine allowing users to conversationally ask questions about the company docs.

Yes, it supports building agents that can use tools, including querying your indexed data, to perform complex tasks. This allows your LLM applications to go beyond simple retrieval and execute multi-step reasoning.

Available index types for structuring data include vector, list, and tree indexes. These structures organize your ingested documents to enable efficient knowledge retrieval and querying for your LLM applications.