Using IQ Bot Process documents action

Use this action to extract values from invoices using a pretrained model.

Procedure

  1. Drag the Process documents action into the workflow.
  2. Use the Learning instance name drop-down menu to select an existing learning instance.
  3. Select one of the following tabs for the input documents:
    • Desktop folder: Enter the folder path or click Browse to choose a folder.
    • Variable: Enter a variable name or browse to select a variable from the list.
  4. Select the folder (instead of a single document) that has all the documents that must be processed.
  5. Select the output folder from the Output folder path field.
  6. Optional: Select the Send documents to IQ Bot Validator for validation check box.
    Although the documents are processed using Bot Runners on your local machine, you must use this option if you want to validate the documents in the IQ Bot Validator.
  7. Optional: In the Group Label field, provide the string variable that holds the name of the folder containing the sorted documents.
    This is necessary for defining a user-created document group. For more information, see Create a new document group.
  8. Click Save.
After the Task Bot is run, the output folder can have up to four subfolders:
  • Success: Contains documents that the package successfully extracted all the fields from and appear as .csv files.
  • Validate: Contains documents that require validation.

    The documents appear as .csv files and IQ Bot extracts the data before you validated it. However, if you select this option, these documents will be available in the Validator and you can download the validated output separately.

  • Retrain: Contains documents that were sent for validation.

    These documents are stored in their original input format and you can train them by uploading them to your existing learning instance. The package does not create new groups. To create new groups and train the failed documents, use the Upload option to send the documents from the Output\Retrain folder to your learning instance. The groups are created based on the Groups Creation Threshold defined in the learning instance.

  • Unprocessed: Contains documents that were not processed due to runtime errors.

    These documents are stored in their original input format and you can process them using the IQ Bot Extraction package.

If you use the IQ Bot Validator, the validated files are stored on the IQ Bot server output folder. Use the IQ Bot > Download all documents IQ Bot: Download all documents action to download files to your output folder on the local device.