The question
How do you replace a legacy form without losing the work it quietly coordinates?
Legacy InfoPath and SharePoint forms—including SCRA, MLA, and account-exit requests—contained conditional rules and permissions that were not always visible on screen. Employees also relied on daily rotations, manual coordination, and knowledge distributed across teams. I studied both the intake experience and the fulfillment process to inform modernization.
My approach
How I worked through it.
Follow the request across roles
I moderated discovery and walkthroughs with operations employees, QC reviewers, analysts, legal and military-servicing advisors. Separate guides captured the needs of people submitting and fulfilling requests.
Recover the hidden specification
AI-assisted analysis of legacy templates helped draft field catalogs, validation logic, role-based permissions, and conditional behavior. I checked those interpretations against source templates and first-hand user accounts.
Make the method repeatable
I organized source material into product-specific repositories and codified a prompt and verification playbook. Research notes also fed import-ready process diagrams for validation with stakeholders.
Evidence to decisions
What mattered. What it changed.
A return queue depended on one person
Single-recipient handoffs could stall when the assigned employee was unavailable.
Design directionDefine shared ownership, reassignment, and routing needs in the future-state requirements.
Missing notifications made work hard to track
Employees described gaps around completed reviews, returns, and cancellations.
Design directionTreat notification events and status transitions as core product behavior.
Second-hand assumptions did not always match actual practice
Interviews revealed conflicting descriptions of submitters, thresholds, and routing.
Design directionMaintain explicit open questions and resolve them with first-hand evidence.
Contribution and outcome
Validated specifications before screen design.
The research delivered structured session analyses, role maps, field-level discovery guides, and prioritized requirement tables. These outputs helped design and engineering understand the operating model before designing a replacement interface.
How to read the outcome
This work delivered research artifacts and requirements. The case study does not include a post-launch evaluation of turnaround time or operational performance.
Reflection
AI was most useful when it made overlooked details easier to inspect. The research value came from validation, reconciliation, and the conversations that followed—not from accepting a generated specification.Let’s talk through the work