Students were not simply struggling to register. Registration friction exposed a broader systems problem: academic workflows had moved into a digital environment without always making process ownership, dependencies, timing, and next steps explicit enough for students or staff to navigate consistently. I used AI to synthesize repeated student-support email patterns into failure types, loops, and system constraints, then applied UX judgment, behavioral science, and institutional knowledge to map where the workflow was actually breaking down.
The Problem
Students most often got stuck when a visible action, like registering for a course, depended on invisible or poorly sequenced conditions: unresolved holds, overrides that addressed only one part of the problem, and unclear ownership across offices. The result was not simple confusion, but a workflow gap, students and staff were trying to complete digital processes that required more explicit dependencies, timing, and routing than the inherited support model made visible.
How I Diagnosed It
I diagnosed the issue by combining AI-assisted email pattern analysis with firsthand systems awareness. AI helped surface repeated failure types and loops, while my UX judgment and institutional knowledge helped explain why those patterns were happening across holds, overrides, seat availability, office ownership, and unclear digital workflow transitions.
What the Patterns Revealed
The patterns revealed that the core problem was not registration itself. Registration was the visible failure point where a deeper workflow issue surfaced.
Students were trying to complete a digital task, but the supporting academic processes had not always been translated into a fully explicit digital workflow. Ownership, dependencies, timing, and next steps often remained implicit, informal, or dependent on local knowledge. That meant students could experience a single task, registering for a course, as a confusing chain of holds, overrides, seat limits, referrals, and unresolved next steps.
AI helped surface the repeated patterns across student-support communication, but the more important finding came from systems analysis: the workflow required explicit process logic that the inherited support model did not consistently provide. In a digital environment, students and staff need visible rules, defined ownership, and clear dependency paths. Without those, the system produces friction even when the people involved are trying to help.