Configure a system data model
Guided data-model setup
Build a reliable model of the records NIM needs before you create filters, mappings, roles, or scheduled jobs.
This guide configures data within one NIM system. Start after the system connection is saved and the connector exposes its available tables. For background on each concept, see Data models.
Build the model in this orderDirect link to Build the model in this order
Choose tables
Select only the object types your workflow needs, such as employees, users, groups, or memberships.
Choose tables to collect →Choose columns
Collect the attributes needed for matching, filtering, relationships, and provisioning, then set their types.
Choose columns to collect →Assign keys
Give each relevant table a stable identifier so NIM can reliably identify its records.
Assign primary keys →Define relations
Connect related tables inside the system, such as employees and departments or users and groups.
Specify relations →What you should have before startingDirect link to What you should have before starting
- A configured system with the connector connection saved.
- A clear workflow outcome, such as provisioning accounts from HR data or assigning group access by department.
- The source attributes needed for that outcome, including a stable identifier for every record type you plan to relate.
- Permission to collect from the system and inspect the resulting records.
Validate after each meaningful changeDirect link to Validate after each meaningful change
Save the model and collect the system whenever you add or change tables, columns, keys, or relations. Then inspect the collected results before building dependent filters or mappings.
| Check | What to confirm |
|---|---|
| Tables | The expected records appear and no unnecessary object types are being collected. |
| Columns | Required attributes are present and have correct data types. |
| Keys | Each keyed table has a stable, unique value for every applicable row. |
| Relations | A sample record reaches the intended related data. |
Common setup sequenceDirect link to Common setup sequence
For an HR-to-directory workflow, you might collect employees, departments, and job_titles; set employee_id, organizational_unit, and job_code as keys; and relate the employee’s department and job-code fields to their respective tables. Once collection confirms the model, create filters to select the employees that need a directory account, then create mappings to provision them.
If a relation or filter does not show expected data, revisit the model before troubleshooting downstream logic. Most issues come from an uncollected column, an incorrect type, an unstable key, or a relation that was not recollected after it changed.
After the model is readyDirect link to After the model is ready
- Create filters to select and shape the collected identity data.
- Configure inter-system relations when your workflow needs to connect records across systems.
- Create mappings to make supported changes in a target system.