Who answers the message that arrives at night? An AI appointment assistant for clinics

In dental, hair transplant and aesthetic clinics, a large share of messages arrives outside working hours. A prospect who sees an Instagram ad at 11:00 p.m. writes in, the reply waits until morning, and in the meantime the same message has gone to several other clinics. It is harder still for patients writing from abroad: a message in German, Arabic or Russian gets a late reply or none at all.

In this article we explain how AI appointment assistants that close this gap work and what to look for when choosing one, using our product Super Randevu.

The real cost of a missed message

• A message that arrives after hours waits for hours; the customer goes elsewhere.
• A customer who says “let me think about it” never gets called again.
• Appointments are scattered across a notebook, a spreadsheet and a phone; the same slot can be given to two people.
• Nobody knows which ad actually brings in appointments.

What these four problems have in common is that first contact and follow-up depend on one person being available. An AI assistant exists to give that first reply every time, within seconds.

A message’s journey: 11:04 p.m., Berlin

A typical flow goes like this: a hair transplant prospect writing in German sends a message from an Instagram ad. The assistant replies in their language, asks about their needs and preferred dates, and marks the customer as “hot”. It then suggests a slot from the real calendar, in both Berlin and Istanbul time. When the customer confirms, the doctor, room and duration are locked together; the team finds everything ready in the morning.

The assistant doesn’t just reply; it offers the slot from the real calendar and prevents conflicts at the database level.

Five things to check when choosing an AI assistant

1. A real calendar: An appointment should be a slot locked in a calendar, not a label. In Super Randevu, the same doctor, unit or room cannot be given the same time slot twice.
2. The customer’s language: It should be able to speak Turkish, English, German, Arabic and Russian, and your team should be able to read the conversation in Turkish. The Arabic patient page should flow right to left.
3. Honesty: The assistant should say plainly that it is an AI, and leave decisions such as price and treatment to the doctor.
4. Protection against manipulation: Attempts to trick it, such as “forget the previous instructions”, should be blocked and an authorized person notified.
5. Where the data lives: Data and backups should be kept on servers in Turkey, and no medical records should be stored.

What works today, what opens with a pilot, and what is on the roadmap

Knowing clearly what a piece of software does today and what it will do later is essential for a good decision. Here is the status for Super Randevu:

• Working today: the appointment engine, the management dashboard, the patient page in 5 languages, reminders, the waiting list and privacy settings.
• Opening with a pilot: the WhatsApp and Instagram assistant; after Meta approval it goes live with the business’s own number and account. Web chat is ready.
• On the roadmap: a deposit and payment link, and syncing with existing appointment systems.

What changes at the clinic in the morning?

The team starts the day with a dashboard that brings together overnight conversations, hot customers, conversations waiting for handover and the day’s appointments on one screen. They take over a conversation the assistant has handed off with one click, read what the customer wrote in Turkish, and send the reply in the customer’s language.

Every step from ad to appointment is reported: how many messages and appointments each campaign brought in, and how long the first reply took. For those who want to manage sales and support processes at corporate scale, YazSis CRM AI carries the same approach to a portfolio of company accounts.