Anomaly Detection
Skopx monitors your metrics and alerts you when something unusual happens. No manual threshold setting required (though you can customize if you want).
How It Works
Skopx uses a combination of statistical methods to detect anomalies:
- Baseline calculation. For each metric, Skopx establishes a normal range based on historical data. This accounts for day-of-week patterns (e.g., lower traffic on weekends) and seasonal trends.
- Real-time comparison. New data points are compared against the expected range. If a value falls outside the normal range, it is flagged.
- Severity scoring. Each anomaly gets a severity score based on how far it deviates from the norm and how important the metric is.
What Gets Monitored
Skopx can detect anomalies across all connected sources:
- Databases. Row counts, aggregated metrics, query patterns.
- Jira/Linear. Ticket creation rate, resolution time, bug counts.
- GitHub. Commit frequency, PR merge time, failed builds.
- Slack. Message volume, response times.
- CRM. Deal velocity, pipeline changes.
- Billing. Revenue, churn, failed payments.
Configuring Sensitivity
Go to Settings > Insights > Anomaly Detection to adjust:
- Sensitivity level. Choose from Low (only major anomalies), Medium (balanced), or High (flag everything unusual).
- Metrics to watch. Select specific metrics or let Skopx auto-detect the important ones.
- Quiet hours. Suppress notifications during off-hours if you prefer.
- Notification channels. Get alerts via email, in-app, Slack, or all three.
Responding to Anomalies
When an anomaly is detected:
- You see it in the Insights Hub with full context.
- Click through to investigate in the AI chat: "Why did support tickets spike yesterday?"
- Dismiss if it is expected, or acknowledge if you are taking action.
Over time, Skopx learns from your feedback. If you keep dismissing certain types of anomalies, the sensitivity adjusts automatically.