Caseload Dashboard

The dashboard at /home/<team>/monitoring is the team's landing page. It is built to be read from the top down.

Data freshness

The banner at the top tells you how current the data is.

Read it first, every time. Everything below it is a statement about a moment in time, and knowing which moment changes what you should conclude. If it is older than you expect, go to Sync and Data Freshness before reading further.

Alerts

Two things sit here.

The alert summary shows which of your configured alerts are currently in violation, and how severely. If nothing is configured yet, this is where you are prompted to set some up.

The alert history chart shows violations over time. This is more useful than the current state: it tells you whether a region has been sitting outside its range for a month or tripped over the line yesterday. A persistent violation is a structural problem; an intermittent one is usually a capacity or seasonality problem.

See Alerts.

Summary cards

Headline numbers for the metrics you are tracking — totals and, where relevant, how they are spread.

Use them as a level check, not a diagnosis. They tell you the size of your operation; they do not tell you whether it is evenly distributed.

Trend charts

One chart per metric, over time.

This is where most of the value is. A single day's numbers tell you very little; the shape over weeks tells you whether a problem is emerging, holding steady, or resolving. Look for:

• Divergence — regions that used to track together separating.
• Sustained direction — steady movement one way over several refreshes.
• Step changes — a sudden jump usually means something changed in the data rather than in the operation. Check the sync history.

The map

An embedded map showing your service area coloured by a metric, so you can see the geographic shape of the distribution without leaving the page.

Use the overlay control to switch which metric is coloured. Some patterns are only visible geographically — a heavy region is one thing, a heavy region wrapped around a light one is a boundary problem.

For the full interactive version, open the Data Editor.

Quick upload

If the team uses a CSV source, the dashboard offers a shortcut to upload a fresh file without going through the Integrations page.

The empty state

A team with no data source sees a prompt to connect one instead of a dashboard. The Data Editor does not appear in the sidebar until a source exists either. See Data Sources.

Reading it well

Start with freshness, then alerts, then trends. Levels last. The absolute number of clients in a region is rarely the interesting fact; how it is changing relative to its neighbours usually is.

Compare regions, not the total. A growing total is a business fact. A total growing entirely in one region is an operational problem.

Distinguish a spike from a trend before acting. One refresh is not a trend, and rebalancing territory in response to a one-week anomaly costs more than it saves.

When something looks wrong, check the data before you check the operation. See Understanding Your Data.

Where to go next