PDF Vector icon

PDF Vector

Convert PDFs, Word, Excel documents, and images to clean markdown, extract structured data with AI, process invoices with specialized parsing, and search millions of academic papers across PubMed, ArXiv, Google Scholar, and more.

Actions18

Invoice → Ask

AI-generated

Overview

This node processes invoices by allowing users to ask specific questions about the invoice content using AI-powered parsing. It supports input via a public URL or binary data from a previous node, and uses different AI model tiers to extract and interpret invoice data. This is useful for automating invoice data extraction, answering queries about invoice details, and integrating invoice processing into workflows.

Use Case Examples

  1. Automatically extract key information from invoices received via email by providing the invoice URL.
  2. Use binary data from scanned invoice images to query specific details like total amount or due date.
  3. Select different AI models based on invoice complexity and file type to optimize accuracy and cost.

Properties

Name Meaning
Input Type How to provide the invoice, either by URL or binary data from a previous node.
Invoice URL Public URL of the invoice file (PDF, DOCX, XLSX, CSV, PNG, JPG). Required if Input Type is URL.
Input Binary Field Name of the binary property from a previous node containing the invoice file. Required if Input Type is Binary Data.
Model AI model tier to use for processing the invoice. Higher tiers produce better results but cost more credits.
Question The question to ask about the invoice. Minimum 4 characters. Required.
Document ID Optional identifier for usage tracking. Returned in the response.

Output

JSON

  • answer - The AI-generated answer to the question asked about the invoice.
  • documentId - The optional identifier provided for usage tracking, returned in the response.
  • rawData - Raw extracted data from the invoice as processed by the AI model.

Dependencies

  • An API key credential for the PDF Vector API

Troubleshooting

  • Ensure the invoice file URL is publicly accessible if using URL input type.
  • Verify the binary data field name matches the output from the previous node if using binary input.
  • Check that the question is at least 4 characters long to avoid validation errors.
  • If the API returns errors related to file size or page limits, select a suitable AI model tier that supports the file size and type.
  • Handle API errors gracefully by enabling 'Continue On Fail' to process multiple items without stopping the workflow.

Links

Discussion