The Extract fields module is a logic module that can extract important information from a text block. This module is particularly useful when extracting key variables from a text to store in a spreadsheet or database.

You can use this module to parse a complex output and extract relevant information for the workflow

This module has multiple configurations:

  • Outputs
    • Each output is one thing you want to take from the input and each one has
      • Field - This is the name of the output for you to see in the workflow
      • Description - This is where you should explain what you want.
      • Eg. Let’s say you want to find the date of when a document was emmited. In Field you write “Date”, and in the Description you put “Date of emission”
  • Extracts as list
    • Enabling will extract values in a list
  • Extract first only
    • Enable if you want the module to only retrieve the first found value

The Extract Information module has one input and variable outputs:

  • Input:
    • Text, receives any text input
  • Output:
    • Text, can have unlimited text outputs; each output requires a proper description as those are leveraged by the AI agent to find the right information.

In this module, you can also use Advanced Options to select the AI model you want to use.

In the image below, the input is a file, which is tranformed to text in the File to Text node, and afterwards, we take the Date (Description: Date of emission) and Company (Description: Company responsible for the document) from the text.