ChatGPT
Claude
Perplexity
Gemini
GrokAn AI agent for university admissions answers, then acts
An AI agent for university admissions answers prospective students' questions from your institution's own content and then takes the next step: it qualifies the applicant, books an open day, records the enquiry in your CRM and hands the conversation to an admissions officer. A chatbot ends at the reply.
In Ireland, the pressure points are the CAO calendar, the Leaving Cert results period and the stream of questions from international students and mature applicants. This guide describes seven concrete use cases, where staff take over and what to check against Irish and EU rules. For the formal definition, see our explainer on the AI admissions agent and its limits.
Use case one: answering course and entry requirement questions
The agent answers questions about courses, CAO points, subject requirements, fees and supports using only approved content, and it says when it does not know. Applicants ask whether a course needs Higher Level Maths, how the Higher Education Access Route works or whether a PLC pathway leads into a degree.
Grounding is what keeps the answers safe. An agent drawing on your course pages and prospectus repeats what you have published and does not improvise. Where the content is silent, it offers a conversation with an officer. Reviewing the unanswered questions each week shows where your website leaves people guessing.
Use case two: qualifying the applicant and routing to the right course
The agent asks a few natural questions, such as entry route, Leaving Cert or mature status, EU or non-EU residency and level of study, and directs the applicant to the right course and team. Qualifying means understanding what someone needs, not scoring them.
A sixth-year student, a mature applicant and a student with a QQI Level 5 or 6 award follow different routes. When the agent separates them early, the record reaching your staff already carries the context. Postgraduate enquiries can go to a department contact instead of the general office.
Use case three: booking open days and campus visits
The agent lists open days, campus tours and information evenings, books the applicant and confirms details. It can also arrange a one-to-one conversation with an access officer or a department.
Booking inside the conversation beats a form followed by emails, because the applicant is still engaged. Families travelling from another county can pick a session that fits their journey. Your events team keeps control of capacity and content, and the agent works only from the sessions you have opened.
Use case four: chasing missing documents and explaining next steps
After an application, the agent explains what happens next and reminds the applicant about outstanding items, such as transcripts, portfolios, interview bookings or evidence of English proficiency. It works from the status your team maintains and never guesses.
Most undergraduate applications go through the Central Applications Office, so the core application sits outside your systems. The agent focuses on what you control: supplementary forms, portfolios, interviews and the steps after an offer. Clear wording about who holds each item prevents applicants from contacting the wrong office, which matters in the busy weeks around the Change of Mind window and the publication of available places.
Use case five: multilingual and out-of-hours answers for international applicants
The agent replies in the applicant's language at any hour, which suits recruitment across time zones. A prospective student abroad can ask about English language requirements or the visa process in their own daytime.
Test the languages you actually recruit in, and test sensitive topics. The agent can point to your published guidance on immigration permission, but it must not interpret the rules or promise an outcome. Those questions belong to your international office. Irish-language queries deserve a check as well if your institution serves Gaeltacht communities.
Use case six: writing the record to the CRM and handing over
When a conversation needs a person, the agent passes it on with a summary so the officer does not start from scratch. It writes the record to the school's CRM: who the applicant is, what they asked, which course interests them and what they consented to.
A strong handover changes the first human conversation. The officer sees the history and the open questions and can begin with something useful. Decide beforehand what triggers it: a request for a person, a complaint, a disability or wellbeing disclosure, or a question the agent cannot answer confidently.
Use case seven: re-engaging a stalled applicant
The agent can follow up with an applicant who asked a question, booked nothing and went quiet, for example by offering an upcoming open day or a reminder about a deadline. These are illustrative workflows, and each depends on the consent given and the channels you may use.
Keep messages brief and relevant. A note that places remain at an information evening is more welcome than a generic prompt. Under GDPR and the ePrivacy rules, direct marketing needs a proper basis, so confirm consent with your data protection officer. If you add WhatsApp, an optional module of Skolbot Chat, check opt-in for that channel first.
The use cases side by side
This table shows what the agent does, where staff take over and what to verify before launch.
| Use case | What the agent does | Where a human takes over | What to check |
|---|---|---|---|
| Course and entry questions | Answers from approved content | Mature entry, exceptions, unclear cases | Is each answer traceable to a page? |
| Qualifying and routing | Asks a few questions, points to a course and team | Complex QQI or credit cases | Are EU and non-EU routes separated? |
| Open days and visits | Lists sessions, books, confirms | Access needs, custom visits | Does capacity match your events system? |
| Documents and next steps | Explains steps, reminds about items | Disputes about what was received | Is status data current? |
| Multilingual, out of hours | Answers in the applicant's language | Visas, SUSI grants, welfare topics | Are tested languages the ones you recruit in? |
| CRM record and handover | Writes the record, passes the chat | Anything needing judgement | Which fields are written, with what consent? |
| Re-engagement | Sends a relevant follow-up | Replies that become real conversations | Is consent recorded per channel? |
What the agent must not do
The agent must not make admission decisions and must not promise anything you have not approved. Decisions belong to your admissions staff, because they carry judgement, fairness and accountability.
The agent should never:
- make or hint at an admission decision, or predict an offer;
- promise a points level, a grant, a scholarship or a visa outcome;
- invent fees, the student contribution, dates, course details or accommodation availability;
- give immigration, legal or financial advice, including on SUSI eligibility;
- continue the conversation when an applicant discloses a welfare concern that needs a person.
Write these rules into the configuration and test them with awkward questions before launch. Ask experienced officers to try to coax the agent into promising a place.
Irish context: GDPR, the DPC, QQI and the EU AI Act
On data, the General Data Protection Regulation applies directly in Ireland, supplemented by the Data Protection Act 2018. The Data Protection Commission is the regulator. You need a lawful basis for processing chat data, a privacy notice reachable from the conversation, a processing agreement with your provider and clarity on where conversations are stored and which sub-processors are involved. Applicants may be under eighteen, so take particular care with consent.
On quality, Quality and Qualifications Ireland oversees quality assurance and the national framework of qualifications, and the Higher Education Authority funds and oversees the sector. What the agent says about programmes and awards must match what you can substantiate, whether you are a traditional university, a Technological University or a private college.
On AI, the EU AI Act sets transparency duties for systems that interact with people. Telling applicants they are talking to an automated assistant is good practice and is easy to build in. Finally, configure the agent for the admissions calendar: the CAO deadline, the Change of Mind window, the Leaving Cert results period and the later publication of available places.
Starting with a thirty-day pilot
Begin with one course area and two actions, and judge the result on your own terms. A narrow pilot lets staff read every conversation.
- Week one: choose a school or faculty, load approved content and agree handover triggers.
- Week two: staff test with common and awkward questions, and you fix content gaps.
- Weeks three and four: go live on selected pages, review conversations weekly and track unanswered questions, handovers and bookings.
Write down beforehand what success looks like, for example fewer unanswered out-of-hours enquiries or cleaner CRM records. For more scenarios, see conversational AI use cases beyond admissions and chatbot scenarios that increase enrolment. The wider picture is in our guide to AI chatbots for student recruitment.
Skolbot Chat is a website agent that answers from your institution's own content, qualifies applicants, books open days and appointments and pushes the record to your CRM. WhatsApp is an optional module.
