ManualTrigger Automate

ManualTrigger Automate streamlines the extraction of data from PDFs by integrating Google Drive and advanced AI models, Claude 3.5 Sonnet and Gemini 2.0 Flash. This workflow allows users to compare results and performance in a single step, enhancing efficiency in data processing while minimizing manual effort. Simply select a document, define your extraction prompt, and receive structured outputs, making it easier to gather critical information like VAT numbers quickly and effectively.

7/8/2025
11 nodes
Medium
manualmediumextractfromfilegoogle drivesticky noteadvancedfilesstorageapiintegration
Categories:
Manual TriggeredMedium Workflow
Integrations:
ExtractFromFileGoogle DriveSticky Note

Target Audience

This workflow is ideal for:
- Data Analysts: Who need to extract specific information from PDF documents efficiently.
- Accountants: Looking to automate the retrieval of VAT numbers from invoices.
- Developers: Interested in integrating AI capabilities with document processing.
- Business Professionals: Who want to streamline their workflow by automating data extraction tasks.

Problem Solved

This workflow addresses the challenge of extracting relevant data from PDF documents without the need for separate OCR and LLM calls. It simplifies the process by combining PDF downloading and data extraction into a single automated workflow, saving time and reducing manual effort.

Workflow Steps

  • Manual Trigger: The workflow begins when the user clicks the 'Test workflow' button.
    2. Define Prompt: A prompt is defined to specify the data to be extracted, in this case, the VAT numbers for each country.
    3. Download PDF: The workflow connects to Google Drive to download the specified PDF document using its file ID.
    4. Extract Data: The PDF is converted to a base64 format, which is required for processing.
    5. Call AI Models: The base64 PDF data is sent to two AI models, Claude 3.5 Sonnet and Gemini 2.0 Flash, for data extraction based on the defined prompt.
    6. Results Comparison: The results from both AI models can be compared to evaluate their performance in extracting the required information.

    • Customization Guide

    Users can customize this workflow by:
    - Modifying the Prompt: Change the text in the 'Define Prompt' node to extract different types of information from the PDF.
    - Selecting Different PDF Files: Update the file ID in the Google Drive node to download different documents.
    - Adjusting AI API Settings: Users can modify the API credentials to use different AI models or adjust parameters such as max_tokens in the API calls for tailored responses.
    - Disabling One AI Call: If users prefer to test only one AI model, they can deactivate the corresponding HTTP request node.