Skip to content
DataMaker Docs

Templates

A template is the unit of test-data generation in DataMaker. You declare a list of fields, each with a type and options, and DataMaker produces rows that match.

Templates are:

  • Versioned — every save is a revision; you can roll back.
  • Reusable — call the same template from the UI, the REST API, the Python SDK, an MCP tool, or a chat prompt. The output is identical.
  • Composable — a field’s type can be “another template”, so you can build complex nested JSON without copy-pasting field lists.
  • Locale-aware — types like First Name, Address, IBAN, and Phone honour a locale setting (de_DE, fr_FR, en_US, …) and produce realistic values.
  • Distribution-aware — fields can be uniformly random, weighted, gaussian, or follow any custom Python.

Anatomy of a template

{
  "id": "tmpl_abc123",
  "name": "Customer",
  "fields": [
    { "name": "first_name", "type": "first_name", "options": { "locale": "de_DE" } },
    { "name": "email",      "type": "email" },
    { "name": "iban",       "type": "iban", "options": { "country": "DE" } },
    { "name": "balance",    "type": "currency", "options": { "min": 0, "max": 10000 } }
  ],
  "version": 7,
  "createdAt": "2026-04-12T10:31:00Z"
}

Every field has a name (used as the JSON key / column name on export) and a type. options is type-specific.

Working with templates

Most teams build templates in the UI — drag-handled field rows, type pickers, options panels, live preview. See your first template.

You can also create or modify templates programmatically:

  • Via the REST API — useful for bootstrapping templates from an OpenAPI spec or a database schema.
  • Via the chat agent — "create a template for SAP Business Partner" reads the OData $metadata, picks types, and saves the template.
  • Via scenarios — programmatically construct field lists in Python.

Sub-pages