Intelligence Engine Version 1.7 Release Notes

Enhancements

Real-Time Supervisor Alerts

Note: Requires Agent Desktop 2.0

Administrators can now configure alerts to immediately notify supervisors when certain intents are detected in live voice and chat interactions. This allows supervisors to immediately intervene at critical moments, such as when a customer asks for a supervisor, a customer reports being in danger, or a call goes silent.

Alerts appear as pop-up notifications, in the navigation panel, and next to the affected agent in the Agents table. Each alert shows the agent, the issue detected, its severity, and whether the interaction is still active.

Supervisors can use the Intent Notification Configuration dialog to select which optional alerts to receive, or limit alerts to their own subteam.

Administrators assign alerts to services and decide which are mandatory for every supervisor.

Alerts are retained for 30 minutes, allowing supervisors to catch up after a break, coaching session, or burst of alerts.

Improved Score Calculation for Negative Weights

We've updated the score calculation formulas for areas and forms that use negative weights to make scoring more intuitive and consistent.

You may notice changes if:

You have areas containing only negative-weight intents. Previously, these areas always scored 0%. Now, they start at 100%, and each detected negative intent reduces the score in proportion to its weight in the area.

You have areas where negative intent weights outweigh positive weights. Previously, these areas could never reach 100%. Now, they can score 100% if all positive intents and no negative intents are detected.

You use areas with negative area weights. Previously, these areas could reduce the overall form score even when no issues were detected. Now, they act as penalties: an area scoring 100% causes no deduction, while an area scoring 0% applies the full penalty.


	Existing scores will not be recalculated. The new formulas apply to newly processed and reprocessed interactions.

< Previous | Next >