Knowledge Graph Builder
An application that transforms unstructured data such as PDFs, documents, text files, YouTube videos and web pages into a structured knowledge graph stored in Neo4j using large language models.
Knowledge Graph Builder lets a user upload files from a local machine, GCS, S3 bucket or web source, choose a preferred LLM, and generate a knowledge graph, extracting nodes, relationships and their properties using a custom or existing schema.
It visualizes graphs for one or several sources at once in Neo4j Bloom, and supports conversational chat with the ingested data, including metadata about the source of a response, through a standalone chat-only route. Supported LLM backends include OpenAI, Gemini, Diffbot, Azure OpenAI, Anthropic, Fireworks, Groq, Amazon Bedrock, Ollama, Deepseek and other OpenAI-compatible base URL models.
It requires Python 3.12 or higher and Neo4j 5.23 or later with APOC installed, since the backend depends on a Cypher variable-scope subquery syntax not available in earlier Neo4j 5.x releases; Neo4j Aura's free tier is supported. It is licensed under Apache-2.0.