How Your Test Suite Can Predict Production Failures with Anomaly Detection

Failed to add items

Sorry, we are unable to add the item because your shopping cart is already at capacity.

Add to basket failed.

Please try again later

Add to wishlist failed.

Please try again later

Remove from wishlist failed.

Please try again later

Adding to library failed

Please try again

Follow podcast failed

Unfollow podcast failed

How Your Test Suite Can Predict Production Failures with Anomaly Detection

Listen for free

View show details

Episode 86 of Software Testing with Fexingo: Lucas and Luna dive into a powerful technique that goes beyond traditional pass/fail assertions: using anomaly detection on test metrics to predict production incidents before they happen. They explore a real-world case where a mid-size e-commerce company used this approach to catch a silent memory leak that standard unit tests missed for weeks. Lucas explains how teams can collect key metrics like response time percentiles, error rates, and resource utilization from their test runs, then apply simple statistical models—like moving averages or standard deviation thresholds—to flag deviations. Luna pushes back on the false positive problem, and Lucas shares how rolling baselines and seasonal adjustment kept the alert noise manageable. The episode also touches on tooling: open-source libraries like PyOD and commercial options like Datadog. Listeners walk away with a concrete playbook for adding a 'prediction layer' to their existing test infrastructure. #SoftwareTesting #QA #AnomalyDetection #ProductionPredictions #TestMetrics #Monitoring #PyOD #DataDog #MachineLearning #MemoryLeak #Ecommerce #DevOps #CI_CD #Reliability #FexingoBusiness #BusinessPodcast #Technology #TestAutomation Keep every episode free: buymeacoffee.com/fexingo

No reviews yet