Neo4j GraphRAG Python library introduces ToolsRetriever class and convert_to_tool() method, enabling dynamic LLM-driven selection of multiple retrievers as tools. This allows combining different retrieval strategies like VectorRetriever for semantic search and Text2CypherRetriever for structured queries, with the LLM intelligently choosing the most appropriate tool(s) for each query. The feature maintains compatibility with existing GraphRAG workflows while providing flexibility for hybrid retrieval approaches.

6m read timeFrom medium.com
Post cover image
Table of contents
Motivation: Flexible and Intelligent RetrievalKey FeaturesGet Oskar Hane’s stories in your inboxImplementation DetailsBenefitsSummaryResources

Sort: