This article lists features, improvements, and bug fixes that are available in sandbox environments only.
Once the features are released to production environments, they are deleted from this article and will appear only in the standard release notes.
Collaboration Room
📂 Agent control over Document Request requirements
Admins can now give agents greater flexibility over how document collection is enforced within a Document Request step. Two new controls are available when enabled by the admin:
Minimum required: Agents can set the minimum number of documents a customer must upload before the step is considered complete — defaulting to all selected documents, but adjustable down to zero.
Step exclusion: Agents can exclude the Document Request step entirely when document collection isn't needed for a particular customer interaction. When excluded, the customer doesn't see the step and the Workflow continues automatically to the next step.
⚙️ Control auto-cropping behavior from System Settings
Admins can now enable or disable auto-cropping directly from System Settings, without needing to contact Customer Support. This makes it easy to adjust the setting if auto-cropping isn't producing the expected results.
Case Manager
🔄 Fix customer contact details without restarting the session
The re-engage feature is now available in customer engagement task. Re-engage enables agents to correct a customer's contact details and resend the collaboration session invitation without losing any existing session content.
Document Analyzer
🛠️ Resolve extraction gaps directly in Document Analyzer
Agents can now edit extracted Attribute values directly in Document Analyzer — including entering a value for an Attribute where no extraction point was found, or marking an extracted value as "missing" when the document does not actually contain that data.
Previously, incomplete or inaccurate extractions could cause rules to fail with no way for agents to correct them in context. With this update, agents can resolve these gaps in real time, reducing rule failures caused by missing or incorrect extraction data and avoiding the need for manual workarounds outside the system.