Skip to content
Systems · Global Incident Early Warning Aggregator

Global Incident Early Warning Aggregator

Documented as a multi-source telemetry pipeline that normalizes signals and runs statistical anomaly detection with confidence intervals for early incident warning.

What it is

The README documents ingestion across four signal types — seismic, epidemiological, climate, and satellite — via HTTP API and streaming (Kafka/NATS), normalized to a [0, 1] scale, then scored by an ensemble of Z-score, IQR, moving-average, and rate-of-change detectors before alert emission.