Document Automation and IQ Bot v.30 release

Review what's new and changed, and the fixes in IQ Bot and Document Automation for the v.30 release.

Note: Information about package updates (such as what's new, changed, fixes, and limitations) is available on package-specific pages. To access these pages, navigate to Package updates overview > <package-name> updates.

For a consolidated list of packages updated in this release, see Packages updated in v.30.

Document Automation


What's new
Data extraction in Document Automation using Generative AI Create learning instances in Document Automation for Cloud with an integrated Generative AI extraction capability that requires minimal configuration. This feature provides improved document processing of unstructured and semi-structured documents with high accuracy. Document Automation data extraction using Generative AI
Integration of third-party parser in Document Automation Using Document Automation SDK, you can now create and connect a custom Automation 360 package as a parser and use it to extract documents. After creating and uploading a custom package to Automation 360, configure the parser with a learning instance and process documents with it.
Custom data manipulation and validation through Document Extraction package (01936705) The Document Extraction package now provides the Get Document Data and Update Document Data actions to allow customization of document fields. With the Get Document Data action, you can retrieve extracted data through a dictionary or JSON file, and use all the available packages to read values and customize them. After manipulating the extracted data, you can use the updated dictionary or JSON string (extracted values or coordinates) as input by using the Update Document Data action and update the documents on the server. This feature helps maximize straight-through processing (STP) and reduce manual verification efforts. Get document data action \| Update document data action
Copy learning instance option added to Document Automation Copy an existing learning instance to reuse the learning instance fields, parser information, document type, language, and local associations in the learning instance copy. Additionally, you can copy the heuristic feedback training and validation rules, as required. Create a learning instance in Document Automation
Notifications for your Document Automation license consumption If you have enabled notifications, you will receive notification for your Document Automation license consumption when the following criteria are met: When 90% of your licenses are consumed When 95% of your licenses are consumed When 99% of your licenses are consumed When 100% of your licenses are consumed This feature allows you to timely plan for additional licenses.


What's changed
Create custom roles and permissions in Document Automation You no longer need to use the system-defined administrator roles for Document Automation roles and permissions. You can create custom roles with limited or no access to administrative permissions while adhering to your organization's security policies. Document Automation users
Enhanced usage of Google Document AI learning instances Add custom fields for Google Document AI learning instances and use them to extract data similar to other existing pre-trained fields. With this enhancement, you can use pre-trained models from Google with custom fields to extract data from documents.
Enhanced language support for Google Custom Document Extractor (CDE) Use Google CDE for any document type in 20 additional languages. With this enhancement, you can: Create learning instances in different languages Add field names Set up Bring Your Own Key (BYOK) license Document Automation support for Google Custom Document Extractor (CDE)
Change in Document Automation license name The Classifier (Number of pages) license is renamed to Document Classifier (Number of pages).
Changes in IQ Bot Classifier package (01955258) The IQ Bot Classifier package has been enhanced to improve its usability: The package name has been changed to Document Classifier package. You can track consumption on the license administration page for processed documents in the Document Classifier package.


Fixes
If you have custom table fields in a learning instance, the table values are now extracted correctly from the specified line item.
If you have used Google CDE for creating a learning instance, the documents are now processed successfully for all the supported languages such as: Albanian, Croatian, Finnish, Icelandic, Slovenian, and Vietnamese.
Document extraction is now successful using proxy in the bring your own key (BYOK) use case.
When you try to create or edit a learning instance, the Create button is no longer disabled if you select the Extract field using pattern check box, and then subsequently clear the check box without entering a value in the Regular Expression field.
When you modify and upload a custom JSON file to the Validator, you can access the file in the Validator without any error messages. Service Cloud Case ID: 01936705
Table names in the Microsoft Form Recognizer Studio interface exceeding 99 characters are now handled without any error messages.
While creating a learning instance, the Create Learning instance button is no longer disabled if you clear the values from any other field.
Use the latest version of the Document Automation package on the previous three (n-3) versions of the Control Room to create a learning instance and process documents successfully. For example, if you are using the Control Room from the Automation 360 v27 version, you can use the Document Automation v30 package to create a learning instance and process the documents successfully.
The name of the downloaded file is now retained even if the name of the learning instance that you have used to process the documents has an empty space.
For the new document level rules, the expected outcome is now reflected correctly in the Validator when the condition is based on a Form Field and the resulting action affects a Table Field.
After you import a learning instance, the Import learning instance from field is now cleared and the imported .dw file is no longer displayed in the field.
When you use the does match regex condition and the show error action to create a validation rule, the fields without any extraction no longer display errors.
If a field name in the Validator has more than 100 characters, the scroll bar appears and you can see the value of the form field. Previously, the form field name was not displayed.
When Extract field using pattern is chosen in the learning instance and a document is processed, the field is validated on entering valid data. Here, only the respective field is affected, and its extraction is based on the provided regular expression input. This was a limitation in builds 18332, 18302, 18277 (Cloud-Sandbox) and build 18324 (On-Premises).
You can now save the documents in French language that are processed through heuristic feedback even when they contain special characters. Service Cloud Case ID: 01995135


Limitations
During validation, when a user clicks on a field, it should open the page in the document with the relevant information for that field or search query. However, for unstructured documents, clicking on a field to validate the value, opens the wrong page.
When you export a learning instance in Firefox browser, you cannot download the .dw file from the download link notification. Workaround: Use Google Chrome or Microsoft Edge-based Chromium to download the .dw files. For more details, see Browser requirements for Automation Workspace.
The message for updating document data in the package does not localize according to the selected language. Additionally, the localized custom message does not translate according to the selected language for Google Document AI support.
The validation user interface shows a stacked view and you cannot change the view.
When a user configures a parser and selects the Japanese language, the Locale field shows the English (united states) value. As a result, user need to select the Japanese locale, if required.
While creating a parser, if a user does not provide the mandatory field values (such as Provider, Document type, and Language), then no error message is displayed to enter mandatory field values. As a result, the user must add all the details to enable the Next button and create a parser.
When a user uploads a JSON file with incorrect parser field values while configuring a parser, the upload process fails and does not show an error message.
When a user edits a learning instance that is created with third-party parser, the Custom aliases option is not disabled.
When a user clicks the Language field during parser configuration, the languages are not displayed in correct order.
During parser configuration, the Configure parser page does not display page content in the selected language.
When a user clicks the Create button while creating a learning instance, an error message displays if you enter a field name with the same case while adding a field.
When a user processes a multi-page document containing header less table with the Google Document AI provider, it extracts only first page from the document.
Sorting does not work when a user clicks the provider name in the Parser Configuration page.
When a user deletes an in between row from validation user interface and try to process same document, it removes the last row from extracted table.
When a user tries to update any field label values while creating or editing a learning instance, it allows you to enter a value of more than 200 characters. After saving the field values, you might encounter an error message for character limit.

IQ Bot


Fixes
Invoices in Japanese language that contain any text in Greek language are now classified correctly by the IDC Classifier and errors are no longer displayed. Service Cloud Case ID: 01967610

Automation 360

Document Automation and IQ Bot v.30 release