Retrieval augmented generation

Document Assistant

View case study
How this demo works

Document Assistant: Your PDF is split into text chunks, searched semantically, and the most relevant context is used to answer your question.

  1. Upload a PDF to create a temporary document session.
  2. Ask a focused question about the uploaded content.
  3. The assistant retrieves relevant passages before generating its answer.

Use non-sensitive files for this public demo. Resetting the demo clears the current client session.

Upload a document to get started.

ℹ️ The file max size is 4.5MB

Then ask things like:

  • → "What is this document about?"
  • → "Summarize the main topics"
  • → "Explain [concept from the doc]"