What is this course about?
Enrollment is where campus AI shows the clearest measurable return and carries the sharpest reputational risk. This course maps the funnel stage by stage, separates volume problems from judgment problems, and treats application reading and financial aid accuracy as the two places where a wrong answer is unrecoverable. It also confronts what CRM vendors will not let you do with your own inquiry data.
Who is this course for?
- VPs and directors of enrollment management
- Admissions directors and application readers
- Financial aid directors
- Marketing and communications staff supporting recruitment
What do I need before starting?
- Familiarity with your CRM and enrollment funnel metrics
- No technical background required
What will I be able to do afterwards?
- Map each funnel stage to whether it is a volume or judgment problem
- Deploy an inquiry-response agent that improves speed without eroding trust
- State the fairness risks in AI-assisted application reading in concrete, testable terms
- Design summer melt outreach with a measurable comparison group
- Evaluate consent and disclosure obligations before an agent initiates contact
What does each module cover?
Which funnel stages are volume problems and which are judgment?
40 minThe diagnostic that determines where AI belongs before any tool is selected.
Objectives
- Map your funnel and classify each stage
- Identify the stages where response speed dominates outcomes
- Identify the stages where a wrong answer is unrecoverable
Topics
Activity. Map your own funnel and mark each stage with its dominant constraint.
How do you answer prospective students instantly without sounding like a bot?
50 minInquiry-response agents, where speed is the whole game and tone determines whether it helps.
Objectives
- Ground an agent in real program, cost, and deadline information
- Design a voice consistent with the institution's brand
- Set the boundary where a prospective student should reach a counselor
Topics
Activity. Build an inquiry agent grounded in three real programs and test it with actual prospect questions.
What are the real fairness risks in AI application reading?
55 minThe module that treats disparate impact as a measurable engineering property, not a talking point.
Objectives
- State disparate impact in terms that can be tested
- Design an audit of AI-assisted reading against historical decisions
- Determine what a reader may delegate and what must remain human
Topics
Activity. Audit AI-generated application summaries against a historical cohort for differential emphasis.
Why is financial aid the highest-stakes accuracy requirement?
50 minNet price and aid eligibility questions, where a confident wrong answer creates a real financial harm.
Objectives
- Identify aid questions an agent must never answer autonomously
- Ground responses in verified institutional aid policy
- Design disclaimers that are honest without being useless
Topics
Activity. Write the aid-question guardrail specification and test it against 30 real questions.
How do you reduce summer melt measurably?
45 minThe clearest return in the funnel — and how to prove the outreach caused it.
Objectives
- Design a melt-prevention outreach sequence
- Construct a comparison group that survives scrutiny
- Identify the operational blockers outreach can actually resolve
Topics
Activity. Design a melt intervention with a pre-registered comparison group and success threshold.
What will your CRM vendor not let you do with your own data?
45 minThe lock-in analysis: what leaves with you, what does not, and what that costs at renewal.
Objectives
- Audit your CRM contract for data portability and model-training terms
- Identify the analyses your current stack structurally prevents
- Compare owned deployment against the incumbent on switching cost
Topics
Activity. Audit your CRM agreement and list every analysis it prevents you from running.
What consent do you need before an agent contacts someone?
40 minConsent, disclosure, and the exposure created when an agent initiates rather than responds.
Objectives
- Determine consent requirements by channel
- Draft disclosure language for agent-initiated contact
- Design opt-out handling that actually works across systems
Topics
Activity. Draft the consent and disclosure package for an outreach agent across three channels.
Building a yield agent on data you own
55 minThe hands-on module: a personalized yield agent grounded in institutional data, not a vendor's shared model.
Objectives
- Deploy a yield agent grounded in your own admit data
- Implement the aid and fairness guardrails from earlier modules
- Establish the measurement plan before launch
Topics
Activity. Deploy the agent against a synthetic admit pool and validate every guardrail before sign-off.
What is the capstone project?
Funnel-wide AI deployment plan with a measurement design
Produce a stage-by-stage plan for AI across your enrollment funnel, including the fairness audit protocol for any application-reading component, the aid-accuracy guardrails, and a pre-registered measurement design for at least one intervention.
Deliverable: A deployment plan an enrollment cabinet could approve and institutional research could evaluate.
How are learners assessed?
- Funnel classification exercise scored against the volume/judgment framework
- Guardrail specification tested against a held-out question set
- Capstone reviewed for measurement rigor as well as operational plausibility
What ships with the course?
Facilitator guide
Session-by-session running order, discussion prompts, and the questions that reliably derail a room.
Learner workbook
Exercises, checklists, and the templates each module's activity produces.
Hands-on lab environment
A sandboxed ibl.ai deployment so exercises run against real agents, not screenshots.
Assessment bank
Scenario questions and rubric criteria mapped to each stated learning outcome.
Source bibliography
Every primary regulation and standard cited on this page, linked and dated.
Which AI agents does this course use?
The hands-on modules run against agents already deployable on the ibl.ai platform for higher education.
Where does the course material come from?
Every module is grounded in primary sources — the regulation, standard, or research itself, not a summary of it. Each was resolved at authoring time.
- Federal Student Aid
U.S. Department of Education
Authoritative source for aid eligibility rules underpinning Module 4.
- Student Privacy Policy Office
U.S. Department of Education
Governs prospective and enrolled student data handling across the funnel.
- AI Risk Management Framework
NIST
Structures the fairness audit methodology in Module 3.
- U.S. Equal Employment Opportunity Commission
EEOC
Reference for disparate impact analysis methodology adapted to admissions.
Delivery notes
Binding guidance for anyone preparing and delivering this course.
- Module 3 is legally and reputationally sensitive. Have counsel review it before delivery, and frame every claim as a testable property rather than a legal conclusion about admissions law.
- Do not use real applicant data anywhere in this course, including anonymized. Build a synthetic admit pool with deliberately embedded demographic correlation so the Module 3 audit finds something.
- Module 6 must not become a competitor teardown. Teach learners to audit their own contract; do not name and attack specific CRM vendors.
- The melt module needs a real worked example with numbers. If no institution will share one, construct one and label it clearly as illustrative.
- Enrollment leaders are outcome-driven; every module should end with what the learner can now measure.
Why run AI training on a platform you own?
You own the course, not a licence to it
Course content, learner data, and the platform run inside your perimeter — you own all the code and the data.
Model-agnostic delivery
Run the course's AI components on any LLM — Claude, GPT, Llama, Gemini, Command — and switch anytime.
No per-seat training licences
Usage-based or self-hosted, so cost tracks actual use rather than headcount.
Deploy anywhere
Cloud, private VPC, on-premise, or fully air-gapped — including for cohorts that cannot use public AI tools.
Frequently asked questions
What does the Enrollment and Yield AI: Agents Across the Funnel course cover?
Enrollment is where campus AI shows the clearest measurable return and carries the sharpest reputational risk. This course maps the funnel stage by stage, separates volume problems from judgment problems, and treats application reading and financial aid accuracy as the two places where a wrong answer is unrecoverable. It also confronts what CRM vendors will not let you do with your own inquiry data. It runs 6 hours across 8 modules across 8 modules, at intermediate level, and closes with a capstone: Funnel-wide AI deployment plan with a measurement design.
Who should take Enrollment and Yield AI: Agents Across the Funnel?
It is written for VPs and directors of enrollment management, Admissions directors and application readers, Financial aid directors, Marketing and communications staff supporting recruitment. Prerequisites: Familiarity with your CRM and enrollment funnel metrics; No technical background required.
Can we run this course on our own infrastructure?
Yes. ibl.ai is model-agnostic and deploy-anywhere — cloud, private VPC, on-premise, or fully air-gapped — and you own all the code and the data. Cohort data, submissions, and any material learners upload stay inside your perimeter, which matters for higher education teams that cannot send work to a public AI tool.
How do we get access to Enrollment and Yield AI: Agents Across the Funnel?
Request access and we will set it up for your cohort — hosted by ibl.ai, or running against your own deployment. Tell us the group size and timing you need, and whether it should run inside your own perimeter.
How much does AI training for higher education cost on ibl.ai?
There is no per-seat pricing — you pay for usage or self-host and pay only for the infrastructure, so a 5,000-person rollout does not cost 5,000 licences. 1.6M+ users across 400+ organizations run the platform this way, including NVIDIA, MIT, and Syracuse University.