Restaurants and cafés · Sample scenario
Reservation assistant
Date, time and party size are extracted from a table request, capacity is checked, and the booking is saved after the customer confirms.
Problem
Reservation messages get late replies: the customer waits hours and goes elsewhere, or the same table is booked twice.
How it works today
The host reads WhatsApp and website messages one by one, checks the booking book or system by hand, writes back to the customer and records the booking manually once they reply. At peak hours messages pile up.
What this scenario changes
AI extracts the date, time and party size from a free-form message. For standard requests, free tables in the reservation system are checked, a suitable option is offered and the booking is created after the customer confirms. Large groups and special events go to a manager.
The process step by step
- Table request: The customer writes on WhatsApp
- Date, time and party size extracted: The message is structured
- Is this a standard booking?: Up to 8 people, no special event
- Capacity checked: Free tables at that time
- Offer drafted: A suitable table is offered to the customer
- Customer confirmation: The customer replies "yes"
- Booking saved: System record and confirmation message
Exception path
- Manager takes the request: Room, menu and terms are agreed
- Customer informed: Told that the manager will get in touch
The role of AI
Extracts the date, time, party size and special requests (birthday, high chair) from a free-form message and writes a short offer to the customer.
Human approval and exceptions
If the party size exceeds the standard limit or the message mentions a special event (banquet, corporate dinner), the system does not confirm on its own; the request goes to a manager with a summary.
Frequently asked questions
What happens if no table is free?
The system offers the nearest free times that day; if the customer declines, the request stays with the host. A table that is not available is never confirmed.
How is the large-group limit set?
The restaurant sets the limit itself and can change it at any time; this scenario uses 8 people as an example.
What if the customer wants to cancel?
The cancellation is received on the same channel, the booking is closed in the system and the table shows as free again.