R2R (RAG to Riches) is a platform designed to help developers build, scale, and manage user-facing Retrieval-Augmented Generation (RAG) applications. It bridges the gap between experimentation and deployment of state-of-the-art RAG applications by offering a complete platform with a containerized RESTful API. The platform includes features like multimodal ingestion, hybrid search, GraphRAG, user and document management, and observability/analytics.
Key Features
- Multimodal Ingestion: Supports a wide range of file types including .txt, .pdf, .json, .png, .mp3, and more.
- Hybrid Search: Combines semantic and keyword search with reciprocal rank fusion for improved relevancy.
- Graph RAG: Automatically extracts relationships and constructs knowledge graphs.
- App Management: Efficient management of documents and users with full authentication.
- Observability: Allows performance analysis and observation of the RAG engine.
- Configurable: Uses intuitive configuration files for application provisioning.
- Application: Includes an open-source React+Next.js app with optional authentication for GUI interaction.
