The manager's view

The manager dashboard reports queue health, worker capacity, channel exceptions, and promise performance. It draws on three reports that refresh every thirty seconds, and each tab stamps the time its figures were read.

Queue coverage

Six headline figures open the screen, each linking to the filtered list behind it: open tasks, active tasks, overdue, service-level breached, open cases, and available capacity.

Per queue, the report carries open, active, overdue and breached task counts, the age of the oldest task, eligible worker capacity, and the number of available workers, with a server-supplied link into the tasks behind each row. Where a queue has work it cannot staff, the gap is reported as one of the four causes above.

A second table reports why work could not be allocated, with a count and the age of the oldest item under each reason.

Queues with no open work and no eligible members are collapsed behind a Show unused queues toggle and, when expanded, labelled as not in use.

Workers

Per worker: kind — human, AI or system — presence state, capacity, active tasks, overdue and breached counts, last activity, and stale accepted work: tasks the worker took and stopped progressing. AI agents appear in the same table with the same accounting.

Interaction exceptions

Missed calls, voicemail, unidentified senders and unlinked messages. Each row carries the channel, direction, status and reason, two matching indicators — borrower matched or unmatched, case linked or not — sensitivity, and age, which is emphasised past sixty minutes. Rows link to the owning task and open the interaction.

Promise and allocation reporting

Windowed over seven, thirty or ninety days. Six figures, then five breakdowns:

There is no per-worker score anywhere on this screen. Decline and cancellation reasons are reported in aggregate by reason code rather than attributed to individuals.