Connecting an API Source

An API connection keeps your caseload data current automatically. Once it is set up, the platform refreshes your metrics every day without anyone touching it.

Start at /home/<team>/integrations and choose the vendor you are connecting. See Vendor Reference for what each one needs.

Before you start

Have the vendor credentials ready — or confirm with your IT team that shared organization credentials are already in place, in which case the wizard will offer them and you will not need the secret yourself.

You will also want to know, roughly, which parts of your service area this team is responsible for. The wizard asks you to confirm that.

The seven steps

1. Credentials

Enter the vendor credentials. The fields differ by vendor. Nothing is contacted yet — this step just collects what the next one needs.

2. Test

The platform makes a real call to the vendor and reports whether it worked.

If it fails, the error is almost always one of: a mistyped key, a key that has been rotated at the vendor, or a base URL or site number that does not match your account. Fix it here — every later step depends on this connection working.

3. Regions

The platform reads the region, branch, or group structure the vendor exposes and shows you what it found.

This is your first real check that you are connected to the right account. If the regions listed are not the ones your team operates, stop and confirm the credentials point at the right instance.

4. Postal Codes

The platform works out which forward sortation areas your service is actually delivered in, and shows them with a client count each, on a map.

Select the areas this team is responsible for. Two things to watch:

• Areas with very small counts are often data-entry errors — a client whose postal code was typed wrong. Including them creates phantom territory.
• Areas you expect but do not see mean either the vendor is not returning those clients, or the postal codes are not populated for them.

You can change this selection later by re-running the wizard.

5. Metrics

Choose which metrics to bring across. A sensible default set is preselected.

Each metric shows a coverage indicator telling you how completely your source actually populates it — a metric that shows as poor or unavailable will produce misleading numbers, so leave it off. Some metrics are shown but disabled because the vendor's API does not expose them. See Metrics Reference.

Fewer, well-populated metrics beat more, patchy ones. You can add metrics later.

6. Preview

A sample of the data as it will be stored: areas, regions, and metric values.

Check the magnitudes. If your organization has around 2,000 active clients and the preview totals 200 or 20,000, something is wrong with the selection you made earlier — go back rather than continuing.

7. Sync

Confirm, and the first sync runs. Depending on how much data your source holds this can take a few minutes.

When it completes, the team's Data Editor appears in the sidebar and the Monitoring dashboard fills in.

After the first sync

Data refreshes automatically each day. See Sync and Data Freshness for the schedule, sync history, and what to do when a sync fails.

Go to the Data Editor next and confirm the regions match how you really operate — the vendor's grouping is a starting point, not necessarily your operational truth.

Changing the connection later

Re-run the wizard from /home/<team>/integrations to change credentials, the selected areas, or the selected metrics.