Release code confidently with Automatic Faulty Deployment Detection

Automatic Faulty Deployment Detection is a feature of Datadog APM that uses machine learning algorithms to spot faulty deployments within minutes and reduce mean time to detection (MTTD). It compares the performance of each new version of a service with its previous versions to identify new types of errors or increases in error rates introduced in a deployment. The tool provides details about affected services, including error types, error rates, request rates, and latency metrics for each deployed version. Automatic Faulty Deployment Detection also enables users to troubleshoot faulty deployments quickly by exploring the service's traces and collaborating with their team through notebooks or incident management processes. Additionally, it allows users to create alerts that automatically page them if a release appears to be faulty, ensuring rapid response to errors and maintaining both development velocity and service quality.