LLM Knowledge Graph Builder — First Release of 2025
Blog post from Neo4j
The LLM Knowledge Graph Builder is an open-source tool developed by Neo4j that enables developers to build retrieval-augmented generation (RAG) experiences for interacting with unstructured data. The tool preprocesses documents, computes text embeddings, and connects them into a lexical graph, which can be used to extract entities and their relationships. This knowledge graph then enables multiple retrievers to fetch data, allowing users to compare the results of different retrievers. The tool also includes features such as community summaries, local entity retrieval, multi-retrievers, retriever evaluation, guided instructions for extraction, user experience improvements, graph consolidation, and more. With its ability to generate context-rich answers and provide explainability, the LLM Knowledge Graph Builder is a valuable resource for developers looking to build more sophisticated AI-powered applications.