Implement Sprint 6.11: Identity & Employee Architecture Cleanup.

Goal:
Finalize separation between login users and employee master data.

Tasks:

1. Create user_employee_map table.
2. Remove dependency on users.employeeId as primary mapping.
3. Make employee import write only to employees and employee_training_targets.
4. Stop auto-creating users during employee import.
5. Update training_records to use employeeId as owner.
6. Keep submittedByUserId / reviewedByUserId for action tracking.
7. Update dashboard, reports, and training matrix to aggregate by employeeId.
8. Keep existing login working.
9. Prepare future Keycloak mapping by empcode.
10. Create docs/sprint-6-11-identity-employee-cleanup-review.md.

Return:

- Migration changes
- Code changes
- Regression checklist
- Review document