Rethinking UX for AI-driven alerting | Datadog
1/22/2019
This post details the evolution of Datadog's alerting UX, moving from traditional static thresholds to AI-driven approaches. It introduces forecasting, anomaly detection, and outlier detection as new methods for generating alerts. Forecasting predicts when a threshold will be reached, anomaly detection identifies deviations from normal behavior, and outlier detection flags dissimilar items within a group. The post also discusses the concept of algorithmic feeds, which can surface important events without explicit user configuration.