The question
What needs to change when the real workflow lives between the tools?
Account-exit specialists worked across disconnected tools, documents, and handoffs. Looking only at the interface would miss the business rules and coordination that made the operation function. I led discovery to connect what employees did, what the systems enforced, and what the future product needed to support.
My approach
How I worked through it.
Observe the work, including the workarounds
I interviewed operations specialists and subject-matter experts, and moderated live walkthroughs with partners in product, engineering, and business analysis. I traced requests across roles rather than treating each screen as an isolated task.
Reconcile the sources
I compared interviews with operating procedures and legacy application logic. An open-questions log distinguished confirmed findings from conflicting accounts, so assumptions could be resolved before they became requirements.
Build a shared view
I mapped 38 process steps across employee actions, touchpoints, backstage activity, systems, and business rules. The blueprint connected friction to future-state requirement tables and product documentation.
Evidence to decisions
What mattered. What it changed.
Coordination was happening outside the system
Employees monitored folders and used off-system communication to keep work moving. The interface did not provide a complete picture of request status.
Design directionMake queues, ownership, and notifications part of the workflow requirements.
Legacy logic was a research source
Undocumented application behavior held rules that interviews alone could not recover reliably.
Design directionUse AI-assisted code and document analysis as a draft, then verify the interpretation with domain experts.
A handoff could be the bottleneck
Mapping multiple roles exposed dependencies and exception paths that a single-user journey would hide.
Design directionSpecify routing and exception handling alongside the primary workflow.
Contribution and outcome
A research foundation for a unified workflow.
The research connected two fragmented account-exit domains in a proposed workflow and supplied traceable input to business and product requirements. It gave the team a shared foundation for modernization, including opportunities to reduce manual quality checks and repeated data entry.
How to read the outcome
The blueprint and requirements are research outputs. Straight-through processing and reduced manual effort are future-state objectives. Post-launch effects have not been measured in this case study. Detailed artifacts can be discussed in a portfolio walkthrough.
Reflection
In enterprise research, the product boundary is rarely the screen boundary. My most useful contribution was making the operating model explicit enough for different teams to challenge, validate, and build from.Let’s talk through the work