# Checking for outliers

Identify outliers and anomalies in your telemetry data using AI SRE's built‑in capabilities.

## Query summaries

When you ask AI SRE to run a [direct query on your logs or metrics](https://betterstack.com/docs/ai-sre/analyzing-data/logs/), use the **Summary** output mode. The AI‑generated summary is specifically designed to highlight anomalies, including:

-   Spikes and drops in values.
-   Changes in error rates.
-   Unusual periodic patterns.
-   Capacity saturation.

The summary always points out "when" the anomaly occurred and "by how much" it deviated from the baseline.

## Chart alerts

Create [alerts](https://betterstack.com/docs/logs/dashboards/alerts/) on your dashboard charts with AI SRE. Choose from three types:

-   **Anomaly alerts**: Uses a Robust Random Cut Forest (RRCF) model to learn the normal shape of your data and fires when current values look unusual. This is ideal for metrics with complex seasonal patterns.
-   **Threshold alerts**: Triggers when a metric crosses a static value. For example, "CPU usage over 90%".
-   **Relative alerts**: Triggers based on a percentage change over a time window. For example, "error rate increased by 50% in the last hour".

This is a proactive way to detect outliers, ensuring future anomalies automatically page the on‑call team.