From FAQ Bot to Service Agent: 24/7 One-Stop and IT Service Desk Agents That Resolve, Not Just Deflect

About this course

AI for student services has been stuck at the FAQ bot: a widget that re-reads the aid office's web pages, answers the question it was asked, then stops. This course teaches the operating model for the next step, a service agent that answers from governed sources, takes bounded actions (file a ticket, check a status, start a request), hands off to a person by design, and is measured on resolution rather than deflection. You will tier one-stop and IT service desk intents as answer, act or route; build a service agent on sample policy and IT knowledge with human handoff; test it with questions that must and must not be handed off; run an evaluation from real transcripts; and cost it per student. Why now: in the 2026 campus CTO survey, 30 percent of technology leaders credit AI with value in IT service management but only 13 percent in advising and support services, while a 2017 randomized trial of a student-facing agent found only 13.5 percent of student messages needed a staff reply. The gap is design, not technology.

What you will be able to do

1. Classify every intent your one-stop and IT service desk handle into answer, act or route, with the source of truth, the owner and the handoff path recorded for each.
2. Configure a service agent that answers only from named, dated institutional sources, declines when it is unsure, and reads a student's own record only under conditions that satisfy the FERPA school-official test.
3. Design at least two bounded actions (a ticket filing and a status lookup) with confirmation, identity and audit requirements written down before they are built.
4. Specify a handoff and escalation matrix with named triggers, ticket states and response-time targets, and an AI-disclosure default that holds regardless of which state law applies.
5. Define and report a metric set that measures resolution, escalation, after-hours share, user rating and downstream outcomes, and explain why deflection alone is the wrong number.
6. Build an evaluation benchmark from real transcripts, run it, and fix three weak spots with evidence from the results.

Who this course is for

Roles: directors of one-stop and enrollment services, registrar and bursar service managers, IT service desk managers and ITSM owners, student-affairs and student-success leaders, and the CIO or digital-experience lead who sponsors a campus-wide service agent.

Institutions: any institution with a front-line service load: large publics with a one-stop center, community colleges with a single student services desk, and private institutions consolidating registrar, aid and IT support into one front door. The course assumes a knowledge base and a ticketing system already exist, in any product.

Before you start: working knowledge of your own service operation (intents, volumes, hours, handoff paths) and of the systems behind it. No technical background is required; the lab uses sample policy pages and a sample IT knowledge base, so no production access is needed.

Format

• Estimated time: 10 to 12 hours.
• Level: intermediate; operational, not technical.
• Hands-on: two labs (build and test a one-stop service agent with human handoff on sample knowledge; build and run an evaluation benchmark from the agent's own transcripts) plus three templates you keep: the intent inventory, the knowledge-source inventory, and the handoff and escalation matrix.
• Structure: 4 sections, each ending in an AI conversation assessment, then a capstone assessment.

Self-paced. Every lesson is taught by an AI tutor that opens with a short story and works through one idea and one question at a time. Each section ends with an AI conversation assessment, and the course closes with a capstone assessment. Section assessments count for 60% and the capstone for 40%; 70% passes. There are no multiple-choice quizzes.

Currency of this course

Last verified 2026-09-28