Rules & Alerts
Configure automated monitoring rules that notify your team when data conditions change, thresholds are crossed, or anomalies are detected.
Creating an Alert Rule
Alert rules continuously evaluate conditions against your data and trigger notifications when those conditions are met. Rules can be created from any chart or data panel using the alert button, or from the dedicated Rules page in workspace settings.
- Navigate to Settings > Rules & Alerts or click the bell icon on any chart.
- Click New Rule and give it a descriptive name.
- Choose the data source and metric to monitor.
- Define the condition type and threshold values.
- Select notification channels and optionally configure escalation.
- Set the evaluation schedule and click Save Rule.
Condition Types
Rules support three classes of conditions, each suited to a different monitoring scenario.
| Type | Description | Example Use Case |
|---|---|---|
| Threshold | Triggers when a metric crosses a fixed numeric boundary (above, below, or equals). | Alert when error rate exceeds 5%. |
| Anomaly | Uses statistical modeling to detect values significantly outside expected range based on historical patterns. | Alert on unusual traffic spikes or drops. |
| Trend change | Triggers when the slope or direction of a metric changes beyond a configured sensitivity. | Alert when weekly revenue growth turns negative. |
Note: Anomaly detection requires at least 14 days of historical data to establish a reliable baseline. Rules configured before sufficient history is available will enter a learning state and begin evaluating once the baseline is ready.
Notification Channels
Configure where alerts are sent when a rule triggers. Multiple channels can be added to a single rule for redundancy.
- Email — Send to one or more email addresses. Supports HTML message templates with metric values and chart snapshots embedded inline.
- Slack — Post to any channel or direct message. Requires the Slack integration to be configured first in the Integrations settings.
- Webhook — Send a JSON payload to any URL. Useful for connecting to PagerDuty, Opsgenie, or custom internal systems.
- In-app notification — Always enabled. Alerts appear in the notification center accessible from the top navigation bar.
// Example webhook payload
{
"rule_id": "rule_abc123",
"rule_name": "Error Rate High",
"triggered_at": "2026-02-27T14:32:00Z",
"condition": "threshold_exceeded",
"metric": "error_rate",
"current_value": 7.4,
"threshold": 5.0,
"dashboard_url": "https://app.openanalyst.com/d/dashboard-id"
}Scheduling and Evaluation
Rules evaluate on a schedule you define. More frequent evaluation increases responsiveness but may consume additional query credits on high-volume data sources.
- Real-time — Evaluates on every data refresh (available on Pro and Max plans).
- Every 5 / 15 / 30 minutes — Suitable for operational metrics.
- Hourly / Daily — Suitable for business KPIs and reporting metrics.
Alert History, Snoozing, and Escalation
The alert history log records every trigger event with a timestamp, the metric value at time of trigger, and the notification status for each channel. Entries are retained for 90 days and are exportable.
Snoozing temporarily suppresses notifications for a rule without disabling it. Set a snooze duration of 1 hour up to 7 days. The rule continues to evaluate and the alert state is still recorded in history; notifications simply are not sent during the snooze window.
Escalation policies define secondary notification channels that activate if the primary channel does not receive an acknowledgement within a configured time window. Escalation is available on Pro and Max plans and is configured per rule in the Advanced section of the rule editor.
Warning: Disabling a rule entirely clears its current alert state. If the condition is still active when the rule is re-enabled, it will re-trigger immediately upon the first evaluation.