Detect anomalies

AI Anomaly Detection scans a time-series metric for statistically unusual data points, so you can quickly tell whether a spike or dip is worth investigating before a meeting.

How to use it

Click Analyze in the chart toolbar and select Detect anomaly. Holistics AI opens and runs the scan automatically. It:

• Looks at the metric's history
• Builds an expected range for each data point from the periods before it (up to 12 months for monthly data, with equivalent windows for other time grains)
• Flags any data points that fall meaningfully outside that range

The result is a chart that shows the expected range alongside the actual values, so you can see why each point was flagged.

Keep monitoring the metric

You can save the generated chart directly to a dashboard to keep monitoring that metric over time.

Find out what drove an anomaly

This pairs with Analyze changes: once you know a movement is statistically unusual, right-click it to dig into exactly what drove it, without leaving your dashboard.

Limitations

• A data point needs at least 3 previous data points before it can be checked. Without enough history, Holistics AI cannot tell whether the change is normal or unusual.
• Seasonality is not taken into account. A dip that happens every year at the same time (for example, a December slowdown) may still be flagged as unusual.