Legal Ease

Legal Ease is a Streamlit prototype for reading a PDF or image, producing a Gemini-assisted summary, and asking follow-up questions about the uploaded document. It is an experiment in making dense legal text easier to inspect, not a substitute for professional legal advice.

What the prototype demonstrates

  • Uploading PDF, PNG, JPG, or JPEG documents
  • Generating a structured summary of parties, obligations, clauses, and risks
  • Asking follow-up questions through a document-aware chat interface
  • Selecting a Gemini model or supplying a separate Gemini API key

The implementation is available in the Legal Ease repository. For a fixed reference, the page also links to the source snapshot used for this project record.

Interface and recorded output

Upload and chat workspace

Legal Ease interface with Gemini model settings, a document uploader, and a follow-up chat field

The prototype places document upload and preview on the left, with the generated summary and follow-up conversation on the right.

Example document summary

Legal Ease displaying a sample loan approval letter beside a structured AI-generated summary

This recorded run shows a sample loan approval letter next to a summary organised around the parties, loan terms, obligations, and risks.

Follow-up question

Legal Ease answering a follow-up question about the consequences of defaulting on the sample loan

After summarisation, the chat keeps the document context so the reader can ask a narrower question about the recorded example.

Demo availability and fallback

The original Streamlit deployment at legal-docs-ease.streamlit.app currently redirects to a Streamlit sign-in page instead of opening the app publicly. The address is recorded here for transparency, but it is not presented as a working proof link.

You can inspect the working evidence without relying on that service: