Anomaly Detection

A refresh succeeds but loads zero rows — the dashboard looks fine until someone notices the numbers are wrong.

Catch data quality issues the refresh status misses.

  • Catch silent data corruption instantly
  • Stop 'successful' refreshes that load no data
  • Know which issues to fix first, automatically

14-day free trial · No credit card · 5-minute setup

Anomaly Detection
4
Active
9
This Week
14
Auto-Resolved
CriticalRow Count Anomaly
FactOrders dropped 42%
WarningMissing Refresh
HR Analytics — 4h overdue
WarningSlow Refresh
Ops Metrics — 186% slower

Why this matters

See the difference anomaly detection makes for your team.

Without SummitView
  • Refresh says 'Success' but data is actually wrong or missing
  • Silent data failures go undetected for hours or days
  • No way to catch anomalies short of manually checking row counts
With SummitView
  • Automatic row count monitoring catches zero-row loads instantly
  • Severity-based triage separates critical drops from normal variation
  • Get alerted before end users see incorrect numbers

See it in action

Here's what anomaly detection looks like inside SummitView.

Anomaly Detection
Tables Monitored
143
Row Count Alerts
3
Stale Data
5
Schema Changes
1
Row Counts
FactOrdersOps Metrics
87,412was 149,803
-41.6%Drop
FactSalesSales Pipeline
1,247,319was 1,244,081
+0.3%Stable
DimCustomerSales Pipeline
44,891was 44,891
0.0%Stable
FactRevenueFinance Daily
0was 891,207
-100%Drop
DimProductInventory Model
2,184was 2,147
+1.7%Stable
Anomaly Detail — FactOrdersDetected 2h ago
Previous
149,803
Current
87,412
Change
-41.6%
13 refreshes agoLatest

Key capabilities

Row Count Monitoring

Detect drops, spikes, and zero-row loads after every refresh.

Duration Anomalies

Catch gradually slowing refreshes before they become timeouts.

Severity Classification

Critical, warning, and info levels for prioritized triage.

Anomaly Timeline

See when issues first appeared and whether they resolved.

What you get in SummitView

Row count anomaly detection with configurable deviation thresholds
Zero-row load flagging as critical anomalies
Duration anomaly detection comparing against rolling historical averages
Missing refresh detection for expected-but-absent refresh events
Severity classification

critical, warning, informational

Anomaly timeline showing when issues first appeared and whether they resolved

Up and running in minutes

Three steps. No code changes. No Power BI modifications.

1

Connect via Cloud Connect or the optional agent — SummitView captures row counts and durations after each refresh.

2

Automated comparisons against historical baselines flag deviations that exceed configured thresholds.

3

Alerts are sent with context — affected dataset, table, deviation magnitude, and recommended investigation steps.

Works with every license tier

No matter your Power BI license, anomaly detection is included.

Power BI Pro

Core monitoring, usage tracking, alerting

Premium Per User

All Pro features + per-table timing, paginated reports

Microsoft Fabric

All PPU features + capacity metrics, throttling detection

Row count anomaly detection requires PPU or Fabric (Enhanced Refresh API). Duration anomaly detection and missing refresh detection work with all license types including Pro.

Catch the failures that 'Success' hides

Start your 14-day free trial. No credit card required. Set up in 5 minutes.

$299/mo per tenant
Unlimited users
All features included