Flexible GraphRAG
Enables hybrid search and AI-powered Q&A by building knowledge graphs from diverse content sources using a configurable architecture.
关于
Flexible GraphRAG is a comprehensive platform designed to streamline document processing, automate knowledge graph construction, and facilitate both RAG (Retrieval Augmented Generation) and GraphRAG capabilities. It offers a highly configurable hybrid search system that combines vector similarity, full-text search (BM25), and knowledge graph traversal across multiple data sources. Built with LlamaIndex, it provides robust backend services via FastAPI and an MCP server, alongside intuitive Angular, React, and Vue UI clients, allowing for seamless ingestion and querying of information.
主要功能
- Automated knowledge graph extraction from documents with LlamaIndex for graph-based reasoning.
- 1 GitHub stars
- Hybrid search combining vector, full-text (BM25), and graph traversal for comprehensive retrieval.
- FastAPI backend with REST API and MCP server providing specialized tools for AI assistant workflows.
- Configurable architecture supporting various vector databases, graph databases, search engines, and LLM providers.
- Multi-source document ingestion from filesystems, CMIS repositories, and Alfresco systems.
使用案例
- Performing comprehensive research and fact-checking across large document corpuses.
- Building and leveraging knowledge graphs for advanced data reasoning and retrieval.
- Generating AI-powered answers to natural language questions based on internal documents.
Sponsored