PORTFOLIO c/o MATTHEW JAY ZUNIGA

VISITOR · --:-- · ·

CASE STUDY 09 OF 14

Agent-based automation concept for a registration workflow

c/o UNC

AI (n.)
— AI tools, apps, adoption work. See: DATA. [VER. 002]

ROLE Graduate Research Assistant, UNC School of Education, Educational Innovation & Technology (EdIT) — follow-up consultant and ideation partnerTIMELINE Aug 2025 – Jul 2026TRACKS ai

WORK IN PROGRESS

OUTCOME Faculty returned repeatedly for help ideating agent-based automation of a registration workflow

What I made

A concept, not a product. While I was a Graduate Research Assistant in UNC's Educational Innovation & Technology group (EdIT), I delivered about 10 one-on-one and small-group trainings for faculty and staff on generative-AI tools. Faculty came back with a further question: could an AI agent automate a registration workflow?

I became the go-to for those follow-up consults, which included ideating agent-based automation for that workflow. This page documents that ideation.

What did the registration workflow register (a course, an event, a program)? How was it run before, which steps did you propose handing to an agent, and which stayed with a person? Has EdIT cleared any of this for public sharing?

Outcome

Faculty returned repeatedly for help ideating agent-based automation of the workflow. That repeat demand is the outcome I can verify. This page describes ideation only; I do not claim a built or deployed system, time saved, or registrations processed.

Did the concept move forward after Jul 2026, and is there a status you are comfortable stating?

Process

The ideation grew out of the follow-up consults that came after the trainings. People seem to think the technical and the creative are separate, but they are at their best together, and that is the point I try to pursue in my technical work.

How did the ideation sessions actually run — what did you and the faculty work through, and in what order? Roughly when did they happen, and how many were there?

Tools

Generative-AI tools were the subject of both the trainings and the ideation.

Which specific tools or agent frameworks did you use or propose during ideation?

NOTESFOR THE PURIST
AUDIENCE
— UNC School of Education faculty who returned after the generative-AI trainings
CONSTRAINT
— Concept stage only; ideated in follow-up consults alongside about 10 trainings and 2–3 hardware, AV, and network fixes per week
TOOLS
— Generative-AI tools
TRACKS
ai

EXIT