Description
Amazon CloudWatch Outlier Detection (often referred to as anomaly detection) applies statistical and machine-learning techniques to CloudWatch metrics to learn expected behavior and highlight deviations. You can visualize expected bands on charts and create alarms that trigger when metrics go outside the learned range. It’s especially useful for noisy, seasonal, or spiky metrics where static thresholds create too many false alerts.
Details
- Pricing model: paid
- License: cloud_service
- API: Available · Docs
Alternatives
More tools from Anomaly Detection & Alerts.
IBM Cloud Pak for AIOps
Enterprise AIOps platform for event intelligence, anomaly detection, and incident automation across tools.
Instana
Application performance monitoring with automated anomaly detection across services, traces, and infrastructure.
Anodot
AI-powered anomaly detection for business and operational metrics with automated root-cause insights.
Chronosphere
Cloud-native observability platform for large-scale metrics with strong alerting and operational insights.
Datadog Anomaly Detection
Observability platform with ML-based anomaly detection for metrics, logs, and application performance signals.
OpenSearch Anomaly Detection
Open-source anomaly detection plugin for time-series data in OpenSearch with alerting integration.