Document Automation and IQ Bot v.31 release

Review what's new and changed, and the fixes and limitations in Document Automation and IQ Bot for the v.31 release.

Document Automation

What's new
Enable or disable external service connection for using generative AI capability in Document Automation Cloud instances based on region of deployment

Administrators can now enable the generative AI capability for Document Automation in the Control Room with a cloud license by navigating to Administration > Settings > Document Automation Settings. You have an option to select the region based on your region of deployment and choose between US and EU.

When this setting is enabled, Professional Developers can create learning instances and process documents using Google Document AI, Google OCR, and Microsoft OpenAI services.

data-extraction-with-gen-ai.html#quj1693429618799__generative-ai-v31 | Configure Settings

Data extraction with generative AI available for invoices, purchase orders, and user-defined document types

We have now extended data extraction with the generative AIcapability to additional domains such as invoices, purchase orders, and user-defined document types. Users can now easily extract data from invoices and custom documents. This feature can be used to query and retrieve data from form fields when used with Automation Anywhere as the extraction provider.

Document Automation data extraction using generative AI
RegexExtract action in validation rules

You can now use the RegexExtract action in validation rules for table and form fields. When a value is extracted from a document, this feature allows users to apply a regular expression (regex) to the extracted value to retrieve a specific substring. It extracts only the first match found in the value based on the regex pattern.

Validation rules in Document Automation | Use regex extract action in validation rule
Advanced training settings option in Document Automation
Using Advanced training settings on the validation page you can now:
  • Adjust or re-map the table fields as required.
  • Set primary column for row identification.
  • Add an end of table indicator value for the system to stop extraction after the value is reached in a table field.
These updates are saved and applied to each document processed subsequently.
Note: To save these settings, you must submit the document.

Use advanced training settings
IQ Bot to Document Automation Bridge package

We now have a new command package that enables you to import all the active IQ Bot learning instances into Document Automation. You can use these IQ Bot learning instances seamlessly in Document Automation and process the documents.

IQ Bot - DA Bridge package

Important: This package is available as a Preview feature for all IQ Bot and Document Automation customers with some limitations. If you encounter any issues, please raise a Support ticket: Open a support case (A-People login required).
What's changed
Extended language support for Google Custom Document Extractor (CDE)

We now support more than 70 languages for document extraction using Google CDE.

Document Automation support for Google Custom Document Extractor (CDE)
Notifications for your Document Automation license consumption

If you have enabled notifications, you will receive a notification for your Document Automation license consumption when 80% of your licenses is consumed.

You can use this feature to plan for additional licenses.

Understanding licenses in Document Automation
Fixes
When a user clicks on the empty space of the document for normalized value for multiple fields, the field values no longer change.

Previously, the value was changed when a user single clicked on the empty space of the document. As a result, the value might be changed accidentally without noticing the behavior.
With improved heuristic feedback model, the extraction now works properly and creates a cluster for certain fields in cases where the previous mappings contain non-English characters, such as Arabic characters or text with single quotes (').

Service Cloud case ID: 01981332, 01981088
The message for updating document data in the package is now localized according to the selected language. Additionally, the localized custom message is translated according to the selected language for the Google Document AI provider.
For unstructured documents in validator page, when a user clicks a field, the correct page is now opened in the document with the relevant information for that field or search query.
When you upload a JSON file with incorrect parserFields values when configuring a parser, the upload process no longer fails, and shows an appropriate error message.
During parser configuration, the Configure parser page now displays page content in the selected language.
When a user processes a multi-page document containing headerless tables with Google Document AI as the provider, all the valid data is now extracted from all the pages.
While creating or editing a learning instance, when a user updates any field label value, the system restricts the user from entering a value not more than 200 characters.

Previously, in such cases, the system allowed the user to enter a value of more than 200 characters, despite the character limitation. And, when the user tried to save the updates, an error message on the character limit was displayed.
No duplicate data is now extracted with heuristic feedback using Google Vision OCR.
Heuristic feedback is now enabled for the Google Document AI provider in On-Premises deployments.
When a user selects a language in the Automation Co-Pilot interface, the auto-fill option label now updates per the selected language during validation.
When a user adds similar alias name as of the field name, the extraction no longer fails. As a result, the document is extracted successfully with specified regex pattern and shown in the validator page.

Previously, in such cases, for example when you added total as alias name for the Total field name, the extraction failed with an error message and you were not able to process documents with regex patterns.

Service Cloud Case ID: 02005661
Limitations
When a user deletes a parser from Parser configuration page and if a user creates a learning instance later, the Document type drop-down list still shows the domain associated with the deleted parser.
The users can create a learning instance when the generative AI feature and the external connection is enabled even if the Cloud license is expired.
You cannot extract feedback based table data if there is no form field.
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.
No error message is displayed to an user when the IQ Bot learning instance has already been bridged by another user.
Limitations from previous releases
When a document rule contains different datatypes in then condition, you might get the wrong error message.

Workaround: If there are multiple fields in the action with different datatypes, structure the document rule such that there is a separate action for each field.
When users with the validator role submit a document for validation, they encounter an error message.

Workaround: You must create a custom role with at least read permission and assign it to the users. For more details about users and custom roles, see Document Automation users and Create a custom role for Document Automation.

Service Cloud case ID: 02092524, 02098383
When user provides a shared network path as output path for Advanced Classifier package actions, the Directory is invalid or not found - null error message is displayed.

Service Cloud Case ID: 02085363
On the Learning instances page, the Google Document AI provider icon is not visible for Google CDE learning instances.
Note: It does not impact the underlying functionality.
When a user applies a formula validation across multiple rows to validate a value at form field level (For example, Sub Total) and if the user deletes one or more rows and resets the value of form field to match with the present rows, the validation error still persists. To fix this issue, perform the following workaround:

Workaround: Users must add zero (0) to the unwanted cell values (quantity, unit price, and total price) and delete the row. Then, add the correct value to the subtotal to clear the error message.
When user process a document for classic Standard Forms learning instance that is connected in Document Automation and if the Bot Agent is installed with proxy authentication, the extraction fails with an error message.

When you process a document using Google CDE with bring your own key (BYOK) setup and the corresponding processor is using foundational model, the document processing fails due to transformational failure.

Workaround: To fix this issue, use the custom model instead of foundational model in Google console.
When you process a document and if it is sent to the validator, you might encounter an issue with decimal numbers (such as .78, .99) for the number data type.

Workaround: To fix this issue, you must enter the decimal numbers as 0.78 or 0.99.

IQ Bot

Fixes
After validation, document that is marked as invalid is shown in the relevant invalid output folder when project ID is undefined in the API and no longer goes missing from the invalid folder.

Previously, users were not able to find and trace the invalid documents after validation and was impacting the document extraction.

Service Cloud case ID: 01960308, 01998792
Limitations
When user connect and process a document with IQ Bot learning instance that contains non-English characters document type (such as Arabic and Chinese characters) and is connected to Document Automation, the license consumption is not tracked in the Control Room > IQBotPages for such connected learning instances.