Datadog Log Anomaly Detection

Automatically Detect Performance Anomalies

Automatically surface anomalies across all of your ingested logs by detecting when a status level sees a substatial increase relative to its historical baseline
Identify spikes and drops in critical health indicators with context-rich alerts that don't require any configuration or setup
Surface concerning patterns and outliers across high-volume log data before they impact your business
Spot faulty canary or blue/green deployments and prevent large-scale failures with automatic code version comparison

Accelerate issue remediation with out-of-the-box, tag-based insights that are surfaced at query time and become more sophisticated as your data set grows
Resolve issues faster by automatically identifying the exact components in your system that are associated with a disproportionate number of errors
Effortlessly uncover latency outliers, such as an exhausted database shard or a view path with degraded Core Web Vitals, in massive volumes of traces and user sessions

Identify the exact users who are experiencing an error in order to quickly assess its scope
Easily prioritize your remediation efforts with automated insight into an issue's impact on frontend views and backend services
Use Impact Analysis data to establish best practices for incident response

Gain instant value from anywhere in the Datadog platform with no additional setup or configuration
Leverage Watchdog’s capabilities to set up customized anomaly, outlier, and forecast alerts for any type of telemetry
Access precise answers based on Watchdog's analysis of your entire environment, including data from 650+ integrations