Case Study: Discipline Tracking & Standardization
Problem
Discipline tracking was inconsistent both within and across factories due to multiple union policies, paper-based processes, and limited HR capacity. This led to delays, inconsistent outcomes, and increased employee grievances.
Users
HR teams and frontline supervisors.
Scale
Initially deployed at a single site, later expanded regionally as part of a standardization effort.
Constraints & Considerations
- Data platform: SharePoint only (Dataverse not available)
- Licensing: No premium connectors or flows (including Word connector)
- Security: Highly sensitive employee data with strict role-based access
- Performance: SharePoint delegation and query limits required careful data access patterns
Solution
I designed and delivered a Power Apps solution optimized for SharePoint’s limitations while maintaining security and performance.
Key decisions included:
- A hybrid, denormalized data model using explicit ID references instead of SharePoint lookups to support more flexible querying and relationships.
- Pseudo-joins performed in-app using targeted gallery logic and temporary collections, carefully scoped to avoid delegation and performance issues.
- A SharePoint-based document templating approach to generate discipline letters without premium connectors. This supported automated document creation, with known limitations around signature handling.
- Role-based access control that allowed HR to manage permissions and reassign incidents while maintaining consistent supervisor roles and restricted visibility.
- Selective use of Power Automate only when necessary, favoring in-app operations for performance unless delegation constraints required a flow-based approach.
Outcome
- Improved consistency and timeliness of discipline actions
- More accurate tracking of policy violations and discipline levels
- Better reporting and visualization of discipline trends
- Employee grievances related to discipline decreased by approximately 10% after launch
What I’d Improve
- While Dataverse would simplify parts of the data model, the more meaningful improvement would be designing the app and data structures to be less site-specific, making it easier to adapt to different union rules and factory requirements.
- A more modular data and UI structure would reduce rework when rolling out to new locations with varying policies.