AI Agent for education centers: from inquiry to enrollment
Enrollment season at a language center has a built-in paradox: the hours when parents are free to ask questions - 9pm to 11pm - are exactly when the admissions team has gone home. Reply the next morning and the parent has already talked to three other centers. That gap is precisely what an AI agent fills best.
An AI agent for an education center does 5 jobs: (1) advises on learning paths and quotes fees by level and goal ("my kid is in grade 5, weak at English, aiming for a selective school"); (2) books placement tests and trial classes straight into the calendar and reminds before the appointment; (3) nurtures hesitant parents - one useful follow-up after 3-5 days with the nearest start date; (4) sends class and tuition reminders to current students; (5) answers parents about curriculum, rules and holiday schedules. All on duty 24/7 across website, Facebook page and Zalo. The golden rule: the agent is transparent about being AI and hands over to a human for complaints or anything specific about a child.
Why education is such a natural fit for AI agents
Three reasons. One, extremely repetitive questions: 80-90% of enrollment-season messages revolve around fees, start dates, entry levels, and Vietnamese versus native teachers - exactly the kind of question an agent answers best. Two, decisions need nurturing: parents rarely commit on first contact; they compare, discuss at home, then... forget. No human team can manually follow up with hundreds of hesitant parents - an agent does it without missing one. Three, a dense appointment schedule: placement tests, trial classes, parent meetings - every no-show is a lost enrollment, and most no-shows are just forgetfulness. These are three of the 9 standard jobs of an AI agent applied to a classroom business.
The 5 jobs - and the line between agent and human
1. Advising and quoting fees. The agent asks 3-4 questions (age, current level, goal, free time slots) and suggests the right course with fees and start dates - instead of making parents decode a price table. 2. Booking placement tests and trial classes. Written straight into the calendar with confirmation, plus reminders 24 hours and 2 hours ahead - no-shows drop visibly once attendance stops depending on memory. 3. Nurturing hesitant parents. Whoever asks and goes quiet gets tagged; 3-5 days later the agent sends exactly one useful message (an open trial seat, a new start date) - never spam. 4. Caring for current students: room-change notices, tuition-due reminders, holiday announcements. 5. Answering parents about curriculum, rules and deferral policies. And the human boundary: complaints, discount negotiations, and any personal story about a specific child - the agent politely hands over with full context, so the advisor never has to ask the parent to repeat themselves.
Before the agent can take a shift, the center must finalize its "textbook": an always-current table of courses, fees and start dates. Stock photo (CC0).
Rollout: where to start, what to measure
Our formula for a mid-sized center: start with one channel (the Facebook page or Zalo OA - wherever parents message most), train the agent on the course table plus 30-50 real questions pulled from the old inbox, run it supervised for 2 weeks, then open it up 24/7. Three numbers to track: the share of after-hours messages answered within 1 minute (before: 0%), the placement-test show-up rate, and how many hesitant parents come back after a nurture message. For the bigger picture of what agents can and cannot do, start with what an AI agent is; for how an education brand goes all-in on AI, see the Duolingo AI-first case study; and before deciding, estimate the payback with our free AI agent ROI calculator - in education, a few extra enrollments a month from the 9-11pm window usually settles the math.
The 5-job framework is distilled from Chạm AI's agent deployments for service and education businesses, 2024-2026; the 80-90% repeated-question figure is an observation from real enrollment-season inboxes and varies by center.
Frequently asked questions
What can an AI agent do for an education center?
Five core jobs: advise on learning paths and quote fees based on the student's level and goals; book placement tests and trial classes straight into the calendar; nurture hesitant parents with scripted follow-ups; send automatic class and payment reminders; and answer parents' questions about curriculum, rules and holiday schedules. All of it runs 24/7 on the website, Facebook page and Zalo.
Do parents mind chatting with an AI?
Not if two rules are followed: the agent clearly introduces itself as the center's AI assistant, and it hands over to a human the moment a parent asks or the question goes beyond its scope - complaints, or anything specific about their child. What genuinely annoys parents is messaging at 9pm and getting an answer the next morning - by then they have already asked three other centers.
Should a small center with 2-3 classes build an AI agent?
Start when enrollment season brings more than 20-30 inquiry messages a day that the team cannot keep up with, or when trial-class no-shows are high because people forget. A single-channel starter package takes about 2 weeks to deploy; below that threshold, a good set of canned answers plus a staff member on duty is enough.