autoflow
AutoFlow: An open-source GraphRAG tool using TiDB Vector, LlamaIndex, and DSPy for building advanced knowledge graph applications.

AutoFlow: Build Knowledge Graphs with GraphRAG
AutoFlow is an open-source GraphRAG (Knowledge Graph RAG) solution that empowers developers to create sophisticated knowledge base tools. It integrates TiDB Vector for robust vector storage, LlamaIndex as its RAG framework, and DSPy for advanced foundation model programming. This combination allows for building powerful conversational search applications and intelligent information retrieval systems, transforming how users interact with complex data and documentation.
Features:
- Conversational Search Interface: Offers a Perplexity-style search page with a built-in website crawler for deep information extraction from specified sites via sitemap URLs.
- Embeddable Web Widget: Provides a JavaScript snippet to easily integrate a conversational search window into existing websites for on-demand assistance.
- GraphRAG Core: Specializes in constructing knowledge graphs for Retrieval Augmented Generation, enhancing contextual accuracy of responses.
- TiDB Vector Integration: Utilizes TiDB Vector for scalable and efficient storage and querying of vector embeddings.
- LlamaIndex Framework: Built upon LlamaIndex to provide a comprehensive RAG pipeline for knowledge base interaction.
- DSPy Powered: Leverages DSPy for programming and optimizing interactions with foundation models, improving response quality.
- Open Source (Apache 2.0): Fully open-source under the Apache License 2.0, inviting community contributions and ensuring transparency.
- Dockerized Deployment: Simplified setup and deployment using Docker Compose with clear resource recommendations.
Summary:
AutoFlow enables the development of advanced knowledge graph-based RAG systems. By leveraging TiDB Vector, LlamaIndex, and DSPy, it provides tools for creating conversational search experiences and embeddable AI assistants. This open-source project is geared towards developers looking to build intelligent, searchable knowledge bases from diverse data sources, with a focus on easy deployment and integration for various applications.

Similar to autoflow:


