Digitizing the pre-delivery inspection workflow

The workflow was entirely paper-based with printed checklists, handwritten Work Authorization Forms (WAFs), approvals moving physically between teams. Data loss, delays, and duplicate entry at every step.

The challenge was coordinating technicians, service managers, and sales managers across mobile and desktop, while designing around physical constraints, high-volume approvals, and role-based dependencies.

I approached PDI as a systems problem. The workflow spans multiple roles, each with different tools, contexts, and goals.

I mapped the full state model across two connected lifecycles.

PDI: Open -> In Progress -> Completed
WAF: Draft -> Service Review -> Sales Review -> Approved or Rejected

Each transition triggers notifications, updates queues, and logs actions. Every state answers who needs to act, what they need to see, and how quickly they must respond.

I built a mid-fidelity end-to-end prototype to validate the workflow across roles. It established core patterns such as three-state checklist inputs, inline issue capture, automatic WAF generation, and centralized approvals.

It also exposed a gap: much of the workflow was based on stakeholder input without full on-site observation, which limited real-world accuracy.

PDI checklists include 15 to 20 items across categories, so I grouped them into collapsible sections with progress indicators.

This reduced cognitive load and let technicians focus on one section at a time. Everything auto-saves, removing the need for explicit save actions.

After submission, I designed a flow that automatically converts failed inspection items into a WAF draft. Vehicle data is pre-filled and line items are generated automatically, removing duplicate data entry between inspection and repair authorization.

Managers often adjust repair estimates before approving, so I designed inline editing within the approval screen.

They can modify values in context, with every change logged for traceability, enabling approve with edits in a single action without back-and-forth with technicians.

Early designs missed nuances in terminology and behavior. A simple mismatch like “Description” versus “Complaints” revealed gaps in understanding technician workflows.

This highlighted the need for deeper on-site validation and closer alignment with real usage patterns.

This project required designing workflows, states, and data movement across teams, not just interfaces. I worked across system logic, role interactions, and constraints to ensure the product could function in real operational environments.

The result was a shift from disconnected paper processes to a structured, scalable workflow system.