Using Onyx (formerly Danswer) with Unstructured for Production RAG Chat With Your Docs
Blog post from Unstructured
Danswer, an open-source AI assistant designed for interacting with enterprise documents, has integrated Unstructured's Serverless API to enhance its document processing capabilities, particularly for files stored in Google Drive. This integration broadens the range of file types Danswer can parse by adding support for 13 additional file types, although it currently excludes Google Drive sheets, slides, and docs. The integration process involves minimal code changes and is demonstrated in a detailed guide and video tutorial. Danswer supports various deployment scales, from local machines to cloud environments, with features like user authentication, role management, and a chat interface that connects to any chosen language model. While this integration marks an initial step in collaboration, further enhancements are anticipated to include ingest pipelining and metadata extraction for improved speed and functionality.