Real-Time Alerts & Alert Rules
Last updated · fulliautomatix 1.0.33 · getafix 3.2.65
When Alerts Enabled is turned on for your company, the Desk's Alerts module (https://desk.<domain>/alerts) watches queue service levels, caller wait times, wrap-up outcome spikes, SIP trunks, recording storage, and sign-in security in real time (evaluated every 10 seconds).
1. Notification Bell & Critical Toasts
- Left Rail Bell: Shows the count of currently open alerts for your scope. The badge turns red whenever at least one Critical alert is open.
- Critical Arrival Toast: When a new Critical alert fires, a red toast notification pops up immediately across whichever Desk screen you are viewing.
2. Managing the Alert Feed (/alerts → Feed)
Each open alert row in the Feed tab shows:
- Severity Chip:
CriticalorWarning. - Alert Description & Subject: For example,
Caller waiting 1:15 on Supportor3 calls closed as "escalate" in 60 min. - Repeat Counter (
×N): If the condition fires again within the rule's cooldown window, it increments×Non the existing alert row rather than flooding the feed with duplicates. - Alert Storm Banner: If 10 or more firings occur within 10 minutes, an amber banner appears above the feed (
Alert storm · n firings in the last 10 minutes folded into m rows · k repeat notifications suppressed) so supervisors can immediately spot a broad incident. - Quick Actions:
Open floor/Open queue/Open reports: Deep-links directly to the affected queue or report.Snooze 15 min: Quiets the alert and removes it from the active feed for 15 minutes.Acknowledge: Marks the alert handled and moves it to the History tab with your name and timestamp.
3. Viewing Alert History (/alerts → History)
The History tab lists all acknowledged and resolved alerts, showing when each episode started, how many times it repeated, and which supervisor acknowledged it.
4. Creating and Testing Alert Rules (/alerts → Rules)
Supervisors can create and manage alert rules directly on the Desk under /alerts → Rules (Company Admins can also manage them in the Admin Console under Settings → Alert Rules).
Click New rule and configure:
Condition (Watch for) | Parameters |
|---|---|
| Service level below | Threshold % (1–100), held for N minutes, on a specific queue or all queues. |
| Abandonment above | Threshold % (1–100), held for N minutes, on a specific queue or all queues. |
| Longest wait above | Wait threshold in seconds (e.g., 60), on a specific queue or all queues. |
| Outcome spike | Fires when N calls are closed with a specific Wrap-up outcome (e.g., escalate) within M minutes on a queue. |
| Trunk registration lost | Fires when a SIP provider trunk goes unreachable/unregistered. |
| Recording storage unreachable | Fires if the recording sidecar (obelix) or object store fails health checks. |
| Failed sign-ins | Fires when N failed login attempts occur within M minutes across the company. |
For every rule, you also choose:
- Severity:
WarningorCritical. - Cooldown (minutes): Minimum time before repeat notifications are sent for the same episode.
- Delivery Channels: In-app (live Desk feed + bell) and/or Email (with a comma-separated list of recipient email addresses).
Rule Controls
Test: Fires a syntheticTest: <Rule Name>alert immediately to verify in-app feed grouping and email delivery.On/Off: Enables or disables the rule without deleting it.Edit/Delete: Updates the rule or deletes it (deleting a rule also clears its historical alerts).