Choose columns to collect
Data-model setup · Step 2
Collect the attributes your workflow needs, give them the right data types, and validate the results before building keys or relations.
Columns are the individual attributes in a selected data table—for example, an employee ID, email address, department, or account status. NIM can use collected columns in filters, relations, mappings, roles, and reports.
Choose columns based on the outcomeDirect link to Choose columns based on the outcome
| Workflow need | Columns commonly needed |
|---|---|
| Identify and match records | Stable system ID, employee ID, immutable directory ID, or another reliable identifier. |
| Provision accounts | Name, email, status, account location, manager, and the source fields required by the target mapping. |
| Assign access by organization | Department, job title, location, cost center, manager, and the corresponding identifiers. |
| Manage groups | User and group identifiers plus the membership attributes that connect them. |
The NIST Privacy Framework is a useful lens for deciding which attributes NIM should collect and who needs access to them. For each sensitive column, record its workflow purpose, access path, and retention expectation before enabling collection. Review those choices when the workflow or source dataset changes.
Select columns in NIMDirect link to Select columns in NIM
- Expand the system’s table list and open the table you want to configure.
- Open the Settings tab.
- In Properties, select each source column to include in collection.
- Select Save.
- Open the Columns tab and refresh the page so the newly selected columns appear.
- Set the appropriate Type for every new collected column.
- Select Save, then repeat for each selected table.
Set column types deliberatelyDirect link to Set column types deliberately
The column type tells NIM how to interpret a value when comparing, filtering, sorting, or using it in a relation. For example, values that contain multiple entries should use the appropriate array type instead of being treated as one plain string.
| Check before saving | Why it matters |
|---|---|
| Does the type match the source value? | A date, number, Boolean, text value, and multi-value attribute should not all be handled as plain text. |
| Is the value single or multi-value? | Membership and entitlement-style attributes may contain several values and need an array type. |
| Will the column become a key or relation value? | Consistent types and formats are essential for dependable matches. |
Missing or incorrect column types can cause filters and relations to return unexpected results or fail. Verify types before moving on to keys and relations.
Connector-specific attributesDirect link to Connector-specific attributes
Some connectors require additional setup before custom attributes appear for collection. For example, Google Workspace custom user-schema fields require the Google Workspace custom schema configuration before they can be collected.
Refer to the relevant integration page for connector-specific table, attribute, and permission guidance.
Validate and continueDirect link to Validate and continue
- Collect and load the system.
- Inspect a sample of collected records and confirm values are present, formatted as expected, and available with the intended type.
- Assign primary keys to the tables that need a stable record identity.
Need a calculated value rather than a source attribute? Create a custom JavaScript column after the source columns it depends on are collected.