We should showcase data drift detection and how we can detect and maybe retrain the model based on updated anomaly data.