For student accommodation
Student accommodation chatbot on WhatsApp, for halls and PBSA
Bedel is an AI chatbot on the WhatsApp number your residents already use, set up for the way each building runs. It answers the questions that come round every week from the building's own manual, turns requests and repairs into tasks with an owner, and puts on the phone of the right person whatever a chatbot should not decide.
What a WhatsApp chatbot does in a student building
A resident writes to the hall or building's usual number. The chatbot works out what they want, answers from the centre's manual when the answer is written down, and acts only where the building has defined what should happen next.
- Answers questions about dining, opening times, house rules, parcels and common rooms, in the resident's language.
- Logs repairs and other issues and assigns them to the right role and shift, with a deadline.
- Handles meal cancellations, packed lunches, bookings of shared rooms and arrival or departure notices.
- Keeps every conversation and every ticket in the team's dashboard, with the time of each step.
What it answers, and what it hands to a person
The rule is short: the chatbot answers information and hands on judgement. The table sums it up; the full version, category by category and with what a good handover looks like, is in the guide to what an AI chatbot in student housing should answer and what it must escalate.
| What comes in | Answers alone | Opens a ticket | Alerts a person |
|---|---|---|---|
| Opening times and house rules | Yes, from the manual | No | Only if the manual does not cover it |
| Meal cancellations and packed lunches | Logs them for the kitchen | No | Allergens: the kitchen confirms |
| A routine repair | Acknowledges it and gives the ticket number | Yes, with its deadline | Maintenance on shift |
| An urgent repair (water, fire, gas) | Only what is essential for safety | Yes, as urgent | The whole on-call post at once |
| Health, conduct, harassment, complaints | No | No | Always, straight away |
Anything not written in the manual goes to a person by default.
What a useful residence chatbot needs
A generic chatbot can answer an FAQ. A student building also needs it to know what it may decide, who should be alerted and when a person must take over.
- Answers anchored to the centre's manual and approved FAQs, given to it separately and with priority, not to whatever a model happens to know.
- Routing by role, named person and shift, without sending every alert to the whole staff group.
- Escalation to a person for money, health, conduct, exceptions or anything the manual does not cover.
- Meta's official WhatsApp Business Platform, so the number is the building's and not someone's personal phone.
Repairs with an owner, a deadline and a record
A photo, a voice note or a line of text is enough. The chatbot opens a repair ticket, classifies it by category and urgency, gives the resident the ticket number and sends it to the person who handles that category on the current shift, on their own WhatsApp, with buttons to accept it and say when they will come.
- Each category has its deadline in hours, which you set. Whoever owns the ticket is reminded at 50 % and at 85 % of the deadline.
- An urgent ticket that runs out rings again once a day, and after three days management is told.
- The resident hears when the ticket is accepted, when it is resolved and if it is reopened.
- Every ticket goes into a spreadsheet with the columns of your own log, and each night management gets the day's summary by email.
Dining, packed lunches and notices
In a hall with a dining room, a large part of the messages are about food. The chatbot logs every cancellation and every packed lunch where the kitchen sees it, so the day's count does not depend on who was at reception when the message came in. In the other direction, the team sends announcements to residents from the same number, and they go out exactly as written.
- "Not in for dinner tonight" and "packed lunch tomorrow, please" go straight into the kitchen's count.
- Announcements can go to the whole building or to one floor; the chatbot does not rewrite or translate them.
- Allergen questions about a specific dish are confirmed by the kitchen, not by the chatbot.
International students, one number
In a building full of international students, language is not a detail: it is half the messages. The chatbot replies in the language of the resident's last message and leaves untouched what must not be translated: times, links and proper names.
- If a reply was about to go out in Spanish to someone who does not write in Spanish, it is translated before it is sent.
- Voice notes are transcribed and treated like text.
- On the staff side, notices, buttons and commands are in Spanish today.
Data in the EU, and the AI Act
The database lives in the European Union, in Ireland; the AI model runs in the US under standard contractual clauses, and its provider does not train on the centres' data. The residence is the controller and Bedel its processor, with the processing agreement published. Under Article 50(1) of the EU AI Act, residents must know they are talking to an AI: the chatbot says so from the first message. More in the guide to the AI Act and a residence's WhatsApp assistant.
Who it is for
It is designed for teams running a hall of residence, a student residence, a PBSA building or a coliving space that want to keep WhatsApp as the front door. The residences that use it today are in Spain; for operators elsewhere, the overview is AI for PBSA and student accommodation.
- Management teams that need to know what is still open at the end of the day.
- Reception, maintenance, dining and housekeeping teams that receive only what belongs to them.
- Operators with several buildings: each building is its own account, with its own rules, rota and deadlines.
Questions about the student accommodation chatbot
Is Bedel a generic chatbot?
No. It is a WhatsApp chatbot set up with each centre's manual, FAQs, roles and shifts. If a question is not covered, it says so and passes it to a person instead of inventing a rule.
Does it replace a hall's management system?
No. Bedel handles the WhatsApp conversation and the day-to-day requests that come in through it; it does not allocate rooms, handle contracts or bill. Those stay in the system you already have.
Does it tell residents they are talking to an AI?
Yes, from the first message. Article 50(1) of the EU AI Act requires it from 2 August 2026, unless it is obvious, and the duty falls mainly on whoever builds the chatbot.
What does it do at night?
It keeps answering what is in the manual. An urgent repair or a safety issue goes to everyone on call at once, at any hour. What can wait is held and delivered when the next shift starts, and the deadline counts from then.
Can a family member get a resident's data?
Not by default. Bedel does not confirm that someone lives at the centre or share their data until the resident authorises it; some actions stay reserved to the resident.
Do residents need to install an app?
No. They write to the building's WhatsApp number, the one they already have. Staff receive their tickets and alerts on their own WhatsApp, with buttons, and the full picture is in the web dashboard.
What happens with a question the manual does not cover?
The chatbot says it does not know and passes the question to the team, rather than making up an answer. The team sees which questions keep coming back without an answer, and adding them to the manual closes the gap for next time.
How can we test it before going live?
On a call, with your own cases: the message at three in the morning, the resident who writes in another language, the fault that keeps coming back. The team can then test replies and alerts in the dashboard's sandbox without messaging real residents.