Agent that summarizes lengthy reports (PDF -> Summary)

How it Works / How to Build It

1. DocumentAPIURL: This node constructs the API URL using the provided document_id from the user inputs.
2. DocumentAPI: This node calls the Vellum API to retrieve the document's contents using the constructed URL and an API key for authorization.
3. ProcessedDocumentURL: This node extracts the URL of the processed document from the API response.
4. ProcessedDocumentContents: This node fetches the actual contents of the processed document using the URL obtained in the previous step.
5. PromptNode: This node generates a summary of the document's contents by prompting a machine learning model with the text retrieved.
6. FinalOutput: This node outputs the generated summary for user display.

What You Can Use This For

• Summarizing lengthy reports for quick insights.
• Creating concise overviews of legal documents for review.
• Generating summaries of research papers for academic purposes.
• Providing quick summaries of internal documentation for team members.

Prerequisites

• Vellum account.
• Access to the Vellum API with a valid API key.
• PDF documents uploaded to the Vellum system.

How to Set It Up

1. Create a new workflow in your Vellum account.
2. Add the DocumentAPIURL node and configure it to accept the document_id input.
3. Connect the DocumentAPI node to the DocumentAPIURL node.
4. Link the ProcessedDocumentURL node to the DocumentAPI node.
5. Connect the ProcessedDocumentContents node to the ProcessedDocumentURL node.
6. Add the PromptNode and link it to the ProcessedDocumentContents node.
7. Finally, connect the FinalOutput node to the PromptNode to display the summary.