CSV Upload

CSV upload is the manual route into the platform. Use it when your scheduling system has no supported API connection, when you are combining data from more than one source, or when you want to explore the platform with a sample.

What the file must look like

One row per geographic area, with the area's region and its metric values.

Required columns

Metric columns

Any number of numeric columns. Name them descriptively — the wizard uses the names to work out what they are.

Example

Region,PC,Active_Clients,Available_PSWs,Visit_Count_per_Week
Region-1,M1L,120,53,369
Region-1,M1N,25,33,93
Region-1,M4A,12,25,43
Region-1,M4B,89,49,434
Region-2,M4C,180,61,622
Region-2,M4J,98,66,420
Region-3,M3C,149,64,729
Region-3,M4G,65,57,153

The rules

1. One row per area. Each postal area appears exactly once.
2. No overlap between regions. Each area belongs to exactly one region.
3. Metrics are numeric. No text, no symbols, no blanks in metric columns.
4. Region names are consistent. North District and North district are two different regions as far as the platform is concerned.
5. There is a header row.

Preparing your export

If your source system holds individual records, you have to aggregate before you upload:

1. Group records by postal area — the first three characters of the postal code.
2. Count clients, workers, visits, and hours per area.
3. Assign each area to exactly one region.
4. Drop every column that is not the region, the area, or a numeric metric.
5. Export as .csv, UTF-8 encoded.

Uploading

Go to /home/<team>/integrations and start a CSV upload. The wizard has five steps.

1. Upload

Choose or drag in your file. It is parsed in your browser first, so you get feedback immediately.

2. Preview

The first rows as the platform read them. Check that columns landed where you expect and that nothing has been shifted by a stray comma or quote.

3. Mapping

Tell the platform which column is the region, which is the geographic identifier, and which are metrics. It proposes a mapping based on your column names — confirm or correct it.

This step is where you also map your column names onto the platform's standard metrics, so that a column called Available_PSWs is understood as active employees.

4. Validate

Two checks run here.

Personal information scan. Headers and sample values are scanned for patterns that look like emails, phone numbers, names, dates of birth, addresses, government identifiers, or card numbers. Anything it finds is shown with the column name and example matches. Treat a high-confidence detection as a stop sign: go back and fix your export.

Data validation. An automated reviewer checks the file for structural and plausibility problems and reports three kinds of finding:

Read the warnings. They are the cheapest chance you get to catch a bad export before it becomes bad numbers on a dashboard.

5. Complete

Review the summary and confirm. Once the import finishes, the Data Editor appears in the sidebar and the Monitoring dashboard populates.

Updating your data

Re-upload a fresh file from the same page. The new upload replaces the existing data — so always export the full picture, every region and every area, not just what has changed.

For active caseload management, weekly or monthly is a reasonable cadence. The Monitoring dashboard shows how old your data is, and the platform will remind you when a CSV source has not been refreshed in a while. See Sync and Data Freshness.

Common problems

More in Troubleshooting.