The KPI alerting that tells you why.

Pulse runs daily anomaly detection on your most-loved metrics, then explains the move in plain English. “Net New ARR is 22% below the trailing-4-week average” is a notification. “Net New ARR is 22% below average, driven by 4 churned customers in the SMB segment, all in the same Salesforce campaign source” is a Pulse alert.

What gets watched

• Any metric defined in the Atlas metric layer.
• Configurable per-metric thresholds: ±X% deviation, absolute floor/ceiling, or 2σ from rolling baseline.
• Automatic seasonality detection (day-of-week + day-of-month).

Where alerts go

Slack (per-channel routing by metric tag), email digests, MS Teams, PagerDuty for severity-1 metrics, and the Pulse home dashboard. Channel rules are managed in the metric YAML, so a new alert on a new metric is one line of config.

How explanations work

When a metric breaches its threshold, Pulse runs an internal contribution analysis: which dimension or segment moved most, which underlying ledger entries are responsible, and which upstream metrics correlated. The model writes the explanation; the source data is linked in the alert. No black-box LLM hallucinations.