What falls outside the norm statistically or semantically — and why that is a common trigger for agents.
Category · Agentics
What deviates from normal.
Anomaly detection identifies data points or patterns that fall outside the expected range, statistically or semantically. In agentic architectures it's a common trigger: an agent kicks off because something doesn't fit the expected picture.
In service triage, predictive maintenance and fraud detection it's a standard component.
See also
