Try it yourself

Run a live analysis, now

Type a metric or dataset name and Agent Forge AI will run a real anomaly + forecast pass against our demo pipeline. Try "checkout_conversion_rate" or a term containing "error" to see a warning result.

zyvenqa-demo-pipeline · sample-events-us● Live
Submit a metric name above to run a live analysis pass.
What just happened

Behind every result

Each analysis pass runs anomaly detection, a 7-day forecast, and a confidence calibration model against the sample pipeline — the same models that run against your real data in production.

Anomaly score

0–1 scale showing how far current behavior deviates from the learned baseline.

7-day forecast

Directional forecast with expected percentage movement over the coming week.

Confidence

Model certainty in its own forecast, calibrated against historical accuracy.