This example shows how the datamart_query_data tool aggregates Datamart data to identify the top customers by revenue. It includes the prompt, request, and response.
Prompt
List top 10 customers by revenue.
Request
{
"uniqueName": "Standard_Sales_Data",
"queryColumns": ["CustomerId", "CustomerName"],
"aggregateColumn": "InvoicePrice",
"aggregateMethod": "SUM",
"top": true,
"limit": 10
}
Response
Here are the top 10 customers by revenue (sum of Invoice Price, all time) from the Standard Sales Data datamart on the copilotfx partition, in EUR:
|
# |
Customer ID |
Customer Name |
Revenue (EUR) |
|---|---|---|---|
|
1 |
CID-0011 |
Stanley Linda CPA FR Ship-to 57 |
10,820,190 |
|
2 |
CID-0057 |
European Distributor DE Ship-to HE |
10,711,177 |
|
3 |
CID-0004 |
Stanley Linda CPA US Ship-to IL |
10,329,548 |
|
4 |
CID-0013 |
Stanley Linda CPA UK Ship-to Hulme |
9,807,386 |
|
5 |
CID-0012 |
Stanley Linda CPA UK Sold-to Westminster |
9,677,245 |
|
6 |
CID-0060 |
European Distributor FR Sold-to 96 |
9,585,152 |
|
7 |
CID-0055 |
European Distributor NL Ship-to DM |
9,484,544 |
|
8 |
CID-0053 |
European Distributor NL Global customer GV |
9,475,331 |
|
9 |
CID-0005 |
Stanley Linda CPA US Ship-to KY |
9,229,814 |
|
10 |
CID-0010 |
Stanley Linda CPA FR Sold-to 8 |
9,206,205 |
The list is tightly clustered (roughly €9.2M–€10.8M) and dominated by two customer groups: Stanley Linda CPA (6 entries) and European Distributor (4 entries). Since these are individual ship-to/sold-to accounts, rolling them up to the parent customer group would show a much bigger gap between the two.
This covers all dates in the datamart. Should I narrow it to a specific period (e.g., the last 12 months) or aggregate by customer group instead?