During upgrade, the installer detects if you have existing learning instances from a prior version of IQ Bot. Select that prior version of IQ Bot from a drop-down list to keep the document classifier version for the existing learning instances.

See Keep learning instance document classifier version during IQ Bot upgrade.

Select between Tesseract4 or ABBYY FineReader Engine FineReader Engine v12.2 as IQ Bot's OCR engine for each new learning instance, and still continue to leverage IQ Bot's native document classification, autocorrection, and extraction capabilities. IQ Bot installation now automatically installs ABBYY FineReader Engine FineReader Engine v12.2, and with an open runtime license. This means that the license is pre-activated, and can be on as many IQ Bot servers you run, and works with or without terminal server.

Access text segmentation and OCR support for 190 languages including: Chinese Simplified, Chinese Traditional, Japanese, Korean

See IQ Bot 11.x: List of languages in IQ Bot.

Log in to IQ Bot without an assigned device license in the Automation Anywhere Enterprise Control Room. The roles are custom or defined by the system. The user view of the UI depends on the assigned role.

See Access IQ Bot without a Bot Creator or Bot Runner device license.

Log in to IQ Bot with a Validator role (defined in Control Room) without a Bot Creator or Bot Runner license. This function supports and permits an unlimited number of Validators in IQ Bot simultaneously.

See Access IQ Bot Validator without a license.

Role-Based Access Control (RBAC) is configured through the Control Room, which enables or restricts access to the IQ Bot learning instances, related features, and functionality that are based on permissions defined in the user role. Create and assign custom roles to users. Without permissions for the custom role, users get an error message when trying to log in to the system.

See Define access to learning instances using custom roles.

The administrator sees action logs for all IQ Bot users in the Audit Logs tab of the Control Room. Successful and unsuccessful actions are logged with reasons for failure logs.

See IQ Bot audit log in Control Room.

The IQ Bot platform administrator is able to enable Windows authentication, during IQ Bot platform installation, to connect SQL databases with Windows or dual authentication.

See IQ Bot Microsoft Windows authentication.

The IQ Bot platform administrator can enable Microsoft Azure SQL database service during IQ Bot platform installation to use the SQL database services.

In the IQ Bot UI, users can now view the number of pages uploaded in production to manually compare with their license limit. Automatic comparison with their license limit is forthcoming in a future release.

Fix field value errors in the IQ Bot Validator. As IQ Bot learns, it can autocorrect field value errors for which it reaches 90% confidence. If users upload additional documents in production, IQ Bot can autocorrect these errors, skip Validator, and count the documents as STP.

See Use Machine Learning to fix extraction errors.

When a user maps check boxes or radio buttons, IQ Bot's artificial intelligence (AI) returns a Yes/No/No check box found value for whether the check boxes have check marks. That applies to a single check box, group of check boxes, and repeated sections of check boxes.

See Extract data from check box or check box groups.

Improved text segmentation, such as, grouping and separation of related or unrelated text, provides more accurate results to support OCR, document classification, and data extraction.

For example, the improvement helped increase STP (percent of documents processed successfully without human intervention) by 9% on a sample dataset of 4,300 documents. The example increase is based on comparing IQ Bot Version 6.5 with IQ Bot Version 6.0.1 on that sample dataset, and was replicated for IQ Bot Version 6.5 versus IQ Bot Version 5.3.0.

See Text segment.

In the IQ Bot Designer, map only one column header per table. If other column headers in the table do not exist or have low quality text, users can map these columns without headers to extract the column data.

In the IQ Bot Designer, for a specific document with repeated tables/sections, users only have to map the first table or section. Automatic extraction occurs for the repeated tables or sections with similar structure or content in the remainder of the document. That automation can continue to production documents.

See Map repeated tables and sections.

In the IQ Bot Designer, select one or more fields to link tables or sections. For example, link health insurance claims data to services data.

That linking automatically continues to production documents, facilitating and eventually merging these tables or sections in RPA.

During document training, a user can look at unique document layouts available in a group, and choose a different document from the default one chosen by IQ Bot Designer, that is more representative of documents across that group.

See Choose training document for a group.

Use formulas for selected fields to flag whether the formulas catch errors. For example, sum of column item totals equals Invoice Total.

See IQ Bot 11.x: Formulas.

When upgrading from IQ Bot Version 5.3.x to 6.5.x, the formula validation also migrates, and this prevent the task of re-adding the formula validation manually.

In the Designer, resize a box around a mapped field value in any direction, without having to redraw or reselect a box. In the Designer, Validator, and Preview mode, (See Extraction Results), hover over any text segment to preview its OCR result.

See Resize mapping area.

In the Designer, Validator, and Preview mode, (See Extraction Results), hover over any text segment to preview its OCR result.

Delete mapping of field labels or field values, by selecting the X delete symbol next to the mapped label or value.

See IQ Bot 11.x: Delete mapping in the Designer.

Select a blue-bounded box to populate the text value for the End of table/section indicator in the IQ Bot Designer.

See Stop extraction at End of table/section indicator.

Stop extraction at End of table/section indicator permits the end of table indicator to work in additional cases.

See Stop extraction at End of table/section indicator.

Option 1: triggers the functionality for Appends only new learning instance(s), if the learning instance ID in the .iqba file (for example, from development environment) differs from the ID in the target environment (for example, production environment). If a learning instance ID in the .iqba file is the same as an ID in the target environment, that .iqba learning instance is not appended.

Option 2: triggers the functionality for Merges new groups and trainings (bots) in existing learning instance(s).

See Using Migration Utility to import and export learning instances.

For international languages, for example, French or German, if a numeric field has a comma misread as a period (100000.00), then an international number pattern (999,00) can autocorrect the values (100000,00).

For the English language, if a numeric field has a period misread as a comma (100000,00), then an English number pattern (999.00) can autocorrect the values (100000.00).

See IQ Bot upgrade options, which includes migration workarounds.

See Import a learning instance.

• Performance and database size are impacted.
• Encryption for learning instance export is not supported for IQ Bot Version 6.5.