The Document Recognition action in Yeeflow is a powerful tool that enables automatic extraction of key information from uploaded documents such as invoices, receipts, and forms. It leverages intelligent recognition technology to identify and parse relevant data, which can then be used to automate tasks, update records, and trigger additional workflow actions. This action integrates seamlessly into Yeeflow's no-code workflow platform and can be customized to fit a wide range of business scenarios.
When to Use the Document Recognition Action
The Document Recognition action is ideal for automating data entry and document processing tasks in a variety of business contexts, including:
Invoice Processing: Automatically extract invoice numbers, dates, vendor names, and amounts from uploaded invoice files to streamline accounts payable workflows.
Expense Management: Recognize and capture key details from receipts and expense forms for faster reimbursement and approval processes.
Form Digitization: Convert scanned paper forms or PDFs into structured data, enabling seamless integration into digital workflows.
Compliance and Record Keeping: Ensure consistent and accurate data extraction for regulatory documents to maintain audit readiness and compliance.
By using this action, you can significantly reduce manual data entry, minimize errors, and accelerate overall processing speed in your document-centric workflows.
Step 1: Setting Up an Azure Document Intelligence Connection
To begin, navigate to Yeeflow's Admin Center. Click the gear icon in the top-right corner, then expand the Integrations tab and select Connections. Next, click on "+ New Connection" and choose Azure Document Intelligence.
A pop-up window will appear where you need to enter a connection name and provide the following details:
API Key: Obtain this from your Azure portal under the Document Intelligence service.
Endpoint URL: The API endpoint for your Azure Document Intelligence service.
Once done, click "Connect" to authenticate and establish the connection.
Configuration Steps
Step 1: Define Action Details
Action Name: Enter a name for the action to identify its purpose within the workflow.
Description: Optionally, add a brief description to explain the action’s functionality or business context.
Step 2: Select Document Source
Select Input File: Choose the file or workflow variable that contains the document to be recognized.
Step 3: Set Service Provider and Connection
Service Provider: Select a document recognition service provider. Currently supported:
Azure AI Document Intelligence
Built-in service
Select Connection（ only available when select Azure AI Document Intelligence) : Choose the appropriate connection configured in your Yeeflow environment
Step 4: Select Processing Model and Page Range
Processing Models: Choose the AI model used to process the document.
Options include:
Prebuilt model: Use a standard model provided by the service.
Custom model: Use a model trained to extract specific fields based on your business needs.
If Custom model is selected, a dropdown will appear to let you choose from your available trained models.
Reload: Click the Reload icon to refresh the list of available models if changes have been made on the service provider side.
Page Range: Define which pages of the document will be processed.
All pages: Select this to analyze the entire document.
Range: Use this option to process only a specific page range.
Step 5: Configure Output and Execution Results
Save Output To: Choose where the processed output data will be stored. This can be a variable or a data field within the workflow for further processing.
Set Values to Variables: Click Select fields to map recognized data fields to specific workflow variables. This allows extracted information—such as invoice number, date, total amount, etc.—to be automatically assigned and used in subsequent steps of the workflow.