AI actions - Extract information from text

Important: 
  • The AI actions feature is available only when the feature is selected in the AI features for the tenant.
  • AI Actions have input limits to ensure safe and predictable results. Each action that analyzes or extracts information from text can process up to 100,000 characters and up to 1,000 repetitive pattern occurrences. Any input that exceeds these limits will not be processed. Some actions that only convert documents or images to text without analyzing the content, do not enforce these limits.

Use the Extract information from text action A tool for building the processes, logic, and direction within workflows. Actions are the steps the workflow performs to complete a process. to extract specific information from unstructured text. The output can be plain text or JSON, based on the details you provide. You can use the result later in the workflow to store values or evaluate conditions.

For example, after a customer submits a written incident report, you may need to extract details such as the claim type, incident date, and location. Use the Extract information from text action to capture these fields and route the result to the appropriate team for assessment and processing.

Configure the Extract information from text

  1. Add the action to the workflow and open the action configuration panel. For more information, see Add an action to the workflow.

  2. Type the text or select a variable that contains the text in the Input text field.

  3. In the Information to extract field, enter the fields or details to extract from the text.

  4. In the Output type field, choose either Text or JSON.

  5. Select or create a variable in the Extracted text field to store the result.

Note: 

AI actions - Extract information from text action fields and settings

Section Field Description

Variable

(Unlabeled) Input text

A variable that contains the text to extract information from.

  • Maximum supported file size is 25 MB.
  • Supported file types include .pdf, .jpg, .jpeg, .png, .bmp, .tiff, .heif, .docx, .xlsx, .pptx , .html

File

  Information to extract The fields or values to extract from the text. Describe using plain language, a list of field names, or a JSON schema. Text
  Output type The format for the extracted result. Choose Text or JSON. Text, JSON
Output Extracted information

Stores the result as unstructured text or structured JSON string.

  • Printed text is supported in over 270 languages.
  • Handwritten text is supported in 12 languages: English, Japanese, Chinese Simplified, Korean, French, Portuguese, German, Spanish, Italian, Russian, Thai, and Arabic.

Object