Alerts

Set up and manage alerts to get notified when important events occur in your project.

Set up alerts to notify you when certain events occur in your project. There are two types of alerts: Event-based alerts that trigger immediately when specific events occur, and Scheduled alerts that run on recurring intervals primarily for observability monitoring.

You’ll need to set up an integration to receive alerts. Connect your Slack, Discord, Teams, or email to receive notifications.

Configure Alerts

Event-Based Alerts

Event-based alerts trigger immediately when specific events occur in your project. These are useful for staying informed about evaluation completions and other discrete events.

To configure event-based alerts:

  1. Navigate to Project SettingsAlerts
  2. Select the Events tab
  3. Toggle on the alerts you want to receive
  4. Click Save

Available Event Alerts

AlertDescription
Notify on Test Run CompleteGet notified every time a test run completes in your project

Scheduled Alerts

Scheduled alerts (also called Custom Alert Rules) notify you on specific recurring events occurring in your project on custom schedules. These are primarily used for observability alerting on traces, spans, and threads.

To create a scheduled alert:

  1. Navigate to Project SettingsAlerts
  2. Select the Scheduled tab
  3. Click New Alert Rule
  4. Configure your alert conditions (e.g., trace count falls below a threshold)
  5. Select your preferred schedule frequency
  6. Click Save

Schedule Frequencies

FrequencyDescription
Every 30 minutesCheck conditions twice per hour
HourlyCheck conditions once per hour
DailyCheck conditions once per day
WeeklyCheck conditions once per week
MonthlyCheck conditions once per month

Example Use Cases

  • Trace volume monitoring — Alert when the number of traces per hour falls below a threshold, indicating potential issues with your application or tracing setup
  • Error rate spikes — Monitor for unusual increases in error rates across your spans
  • Performance degradation — Track latency metrics and get notified when they exceed acceptable thresholds