An AI admissions assistant is an AI agent that answers prospective students’ questions about programmes, fees, deadlines and requirements on the channels they already use, usually WhatsApp and the website, then qualifies each enquiry, sends the right application link, reminds applicants of missing steps and hands complex cases to an admissions officer. It works best on the high-volume, repetitive questions that bury admissions teams in peak season.
This guide is for universities, private training institutes and language centres in Saudi Arabia, the UAE and the wider Gulf. It covers what the evidence says, what to automate, what to keep human and how to set it up. To see how NourSky builds this for the sector, visit NourSky for education.
Why does response speed decide enrolment?
Prospective students shop around, and they judge an institution by its first reply.
- In Ruffalo Noel Levitz’s 2021 Graduate Student Recruitment Report, based on more than 1,500 prospective graduate students, about one in four expected a personal reply (not an auto-responder) “within minutes.” More than 80% said they were likely to enrol at the institution that sent them an admission offer first.
- In Keystone Education Group’s 2023 survey of 23,800 students across more than 195 countries, 62% expected a reply from a university within 24 hours or less, as reported by ICEF Monitor.
- In a 2023 mystery-shopping study of 128 universities by Edified, one in four student enquiries received no response at all.
The competition is local, too. Saudi Arabia’s Technical and Vocational Training Corporation reported 2,234 licensed private training establishments in 2024, with 339 new licences issued that year. A student asking about an IELTS or accounting course can message several of them in minutes.
Elissa Newall, senior partner at Edified, summed it up in a 2026 interview with ICEF Monitor: “The enquiry stage is an incredibly influential moment in a student’s journey.”
Does an AI admissions assistant actually increase enrolment?
The best published evidence comes from a randomised controlled trial at Georgia State University. Lindsay C. Page and Hunter Gehlbach reported the results in AERA Open in 2017. The study included 7,489 admitted students, split randomly between a group that received text messages from an AI assistant called “Pounce” and a control group.
The findings:
- In the authors’ words, students committed to the university who were assigned to the assistant “exhibited greater success with pre-enrollment requirements and were 3.3 percentage points more likely to enroll on time.”
- “Summer melt”, admitted students who never show up, fell by about 21% among those committed students.
- When Pounce couldn’t answer, which happened in 13.5% of cases, the question was routed to admissions counsellors by email. Their answer went back to the student by text and was added to the knowledge base.
Two cautions. First, this is one university in the US, with a specific student population; results elsewhere will vary. Second, the design detail matters more than the headline: the assistant worked because it pushed students through concrete pre-enrolment requirements and because humans answered what the machine couldn’t.
What should an AI admissions assistant handle, and what stays human?
| Task | AI assistant | Admissions team |
|---|---|---|
| Programme details, fees, dates, entry requirements | Answers from approved programme sheets | Keeps the sheets current |
| Qualifying the enquiry (level, programme, start date, sponsor) | Asks and records | Reviews edge cases |
| Sending application and payment links | Yes | — |
| Reminding about missing documents and deadlines | Yes, with approved templates | — |
| Scholarship, discount or exception requests | Records and routes | Decides |
| Visa, equivalency or accreditation edge cases | Routes | Answers |
| Parent concerns, complaints, wellbeing issues | Routes immediately | Handles |
| Corporate or group training enquiries | Collects requirements | Prices and negotiates |
How do you set up an AI admissions assistant, step by step?
- Write a programme sheet for every course. Dates, duration, fees, requirements, accreditation, refund policy, delivery mode. This is the assistant’s only source of truth. If it isn’t on the sheet, the assistant routes the question instead of guessing.
- Map the enrolment pipeline. For example: new enquiry → qualified → application sent → documents complete → paid → enrolled, plus a “next intake” stage for students who aren’t ready yet.
- Connect the channels. WhatsApp, website chat and forms should feed one CRM. On the WhatsApp Business Platform, Meta’s documentation explains that a 24-hour customer service window opens when the student messages you; after it closes, only pre-approved templates can be sent. Deadline reminders a week later must be written as templates in advance.
- Define handoff rules. Name who receives which type of case, how they’re notified, and what the student is told while waiting.
- Design the nudges. Georgia State’s result came from proactive reminders tied to real deadlines, not just from answering questions. Map each pipeline stage to the reminder that moves a student to the next one.
- Tell students they’re talking to an AI. Say so in the first message and always offer a route to a person.
- Review conversations weekly. Every question the assistant routed is a candidate for a new line in a programme sheet.
What does it cost to run on WhatsApp?
Meta moved the WhatsApp Business Platform to per-message pricing on 1 July 2025. According to Meta’s pricing page, non-template messages inside an open customer service window are free, as are utility templates sent within an open window. Templates sent outside the window are charged by category and country. The practical upshot: answering students quickly is cheaper than chasing them days later.
What are the common mistakes?
- Automating before the content exists. An assistant trained on a marketing brochure will confidently repeat outdated fees.
- No human behind it. Georgia State’s assistant routed 13.5% of questions to staff. Budget time for that queue.
- Answering but not tracking. If enquiries don’t land in a CRM with a stage, you’ll never know which programme converts and where students drop off.
- Treating “not now” as lost. A student who wants the next intake is a future enrolment; record the date and remind them when applications open.
For a plain-language definition of the category, read What Is an AI Employee? A Practical Definition for Business Owners. To estimate how many enquiries slow replies may be costing you, try the lost-leads calculator.
See an AI admissions assistant answer your real enquiries
NourSky builds AI admissions assistants for schools, universities and training institutes in Saudi Arabia and the UAE: answers from your programme sheets, qualification and reminders on WhatsApp, a CRM that tracks every applicant to enrolment, and clear handoff to your admissions team.
FAQ
Can an AI admissions assistant answer in Arabic and English?
It can be built to handle both. Quality depends on the approved content you give it in each language, so test it on real student questions before launch.
Will an AI admissions assistant replace our admissions officers?
No. It takes the first reply, repetitive questions and reminders. Scholarships, exceptions, complex eligibility and anxious parents stay with your team, and the Georgia State study relied on staff answering what the assistant couldn’t.
How quickly should we reply to a student enquiry?
As fast as possible. In RNL’s 2021 research about a quarter of prospective graduate students expected a personal reply within minutes, and in Keystone’s 2023 survey 62% expected one within 24 hours.
