Document Automation v.33 release

Review what's new and changed, and the fixes and limitations in Document Automation and IQ Bot for the v.33 release.

Document Automation

What's new
Introducing Anthropic models on AWS and GCP for data extraction in Document Automation

You can use the Anthropic generative AI models that are available via Amazon Web Services (AWS) and Google Cloud Platform (GCP) for data extraction in Document Automation. This offering provides you the flexibility to select the generative AI model depending on the Cloud provider your company has certified.

Anthropic provides the following advantages:

  • Efficient processing of large, unstructured documents
  • Handle documents in both English and other languages
  • Faster processing of documents with better data extraction accuracy
  • Flexibility to use your own license via the bring your own license (BYOL) model

Create a learning instance in Document Automation | Extract data action | Extract data using Anthropic models
Process unstructured documents in more languages

You can now process unstructured documents that include content in the following languages: Afrikaans, Albanian, Chinese (Simplified), Chinese (Traditional), Croatian, Czech, Danish, Dutch, Estonian, Finnish, French, German, Hungarian, Icelandic, Indonesian, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Slovenian, Spanish, Swedish, and Tagalog.

Language support matrix and OCR engines
Process Standard Forms in more languages

You can now process Standard Forms that include content in the following languages: Afrikaans, Albanian, Chinese (Simplified), Chinese (Traditional), Croatian, Czech, Danish, Estonian, Finnish, Hungarian, Icelandic, Indonesian, Malay, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Slovenian, Swedish, and Tagalog.

Language support matrix and OCR engines
What's changed
Improved advanced training settings in validator

When you provide validation feedback for the Primary Column and the End of table indicator fields, the validation feedback is now improved to automatically populate these values when you process documents of similar type even when the end of table indicator is at different locations.

You can clear the automatically updated values:

  • Primary column field: Click the drop-down menu and select the empty value from the drop-down menu.
  • End of table indicator field: Click the close button next to the value in the end of table indicator field or click the close button on the selection box of the value on the document.

Improve table data extraction
Fixes

You can now use field rules that involve multiple replace and regex actions on the same field, or use multiple field rules with replace and regex actions on the same field.

Previously, you were unable to use field rules in such scenarios.
Data extraction now works in debug mode for learning instances that were moved from IQ Bot classic to Document Automation.

Previously, data extraction failed in such a scenario.

Service Cloud Case ID: 02138473
You can now process documents in Document Automation using a migrated classic IQ Bot learning instance when the learning instance was created using check box.

Previously, data extraction failed in such a scenario.
When a user clicks the Language field during parser configuration, the languages are now displayed in the correct order.

Previously, the languages were not displayed in the correct order in such a scenario.
You can now provide a shared network path as the output path for the Advanced Classifier actions and successfully execute bots using these actions.

Previously, an error was displayed in such a scenario.

Service Cloud Case ID: 02085363
When you validate documents, you will no longer see an error when you navigate from the Detail view to the Table view.

Previously, an error was displayed in such a scenario.
When a user processes a document on a custom process learning instance and submits the document, the Validate documents count for the instance is updated correctly.

Previously, the count was not updated when the document was processed, and a negative count was displayed when the document was submitted.
You can now view the Google Document AI provider icon when you use the Receipts and Utility Bill document types.

Previously, the provider icon was not displayed in such a scenario.
You can now process documents that contain decimal numbers (such as .78, .99) for the number data type in the validator.

Previously, an error was displayed in such a scenario.
Limitations
When the Generative AI driven data extraction option is enabled and you select a generative AI provider (OpenAI or Anthropic) when creating a learning instance, you will not be able to change the provider later. The option to switch to a different generative AI provider will not be displayed when editing the learning instance details.
When you use the Connect Learning Instances from IQ Bot feature in Document Automation and if the number of learning instances is more, you will be unable to use the scroll bar to view all the learning instances.

Workaround: Use the mouse scroll button to scroll through the learning instances or use the search option to view a specific learning instance.
If you are using the Connect Learning Instances from IQ Bot feature in Document Automation to connect to learning instances in IQ Bot, if documents processed using this option are queued for validation, and if you update to v.33 release, you might see a blank page when you try to validate the documents that were queued in certain scenarios.

Workaround: Click View status and then check the pending tasks that are queued for validation.
If you copy a learning instance that is using a third-party parser configured in Document Automation and process documents using the copied learning instance, data extraction will fail.
A user with the Automation Co-Pilot administrator permissions is unable to view the Document Automation tasks that are assigned or requested and are in pending or complete status.
If a document contains multiple values similar to the value submitted as the end-of-table indicator during validation (for example, total, subtotal), the system will use the first occurrence of the value as the end-of-table indicator. This value might not always be the value submitted as the end-of-table indicator during validation when you reprocess the document or documents of similar formats.
When you set all the table fields in a learning instance to optional and if any documents processed using this learning instance are sent for validation, the validator will not display an option to add a row for the table fields.

Workaround: Set at least one of the table fields to required or delete the empty row from the table to see the Add row button.
When you use the IQ Bot - DA Bridge package and the Connect Learning Instances from IQ Bot feature simultaneously and if you delete, re-import, or reconnect any learning instance, you might see an incorrect user interface for the learning instance.

When extracting data using a generative AI provider, a specific field might return an empty value if the response is requested in JSON format within the search query.

Workaround: Instead of requesting the field to return the query in JSON format within the search query, request the query result to be in a different format, such as an array or text.
When you are using the IQ Bot Pre-processor package actions and if the Input file path and Output folder path contains Japanese characters, you will see an error when processing documents.
Workaround:
  • Place the documents to be processed in a folder path that does not contain any Japanese characters and provide the path in the input file path field.
  • Create an output folder in a folder path that does not contain any Japanese characters and provide the path in the output folder path field.
Limitations from previous releases
No error message is displayed to an user when the IQ Bot learning instance has already been bridged by another user.
For a public process, you might encounter an error message in the following scenarios:
  • After validating all the documents in the validation queue.
  • After processing some documents with same learning instance, if you open the first document and click Refresh.
Note: No error message is displayed for the private process.

IQ Bot

Fixes
When using IQ Bot for extracting large amounts of data that cannot be stored on your local (browser) storage, you will no longer see the “Something went wrong” error when validating documents.

Previously, you were seeing this error in certain scenarios.

Service Cloud Case ID: 02094667

Updates to the interface

Document Automation
When you create a learning instance using Generative AI-driven data extraction, you have the option to select from OpenAI or Anthropic LLM option in the UI.

Image demonstrates the new option for selecting LLM model.

See New LLM model options.
When you select the end-of-table indicator during document validation, you can use the new close icon to clear the end-of-table indicator selection.

Image demonstrates the close option.

See New end-of-table indicator icon.