AI receptionist and chatbot for clinics
An AI receptionist for clinics answers patients on WhatsApp, on the website and by phone, handles the usual questions and logs the appointment request or message in the CRM. It only answers with the information the clinic gives it, says it is an AI and passes anything medical to a person. We build it at Elevatec Techlearn, an artificial intelligence agency and fractional CTO based in Madrid.
Who it is for
It makes sense in clinics where the phone and WhatsApp take up a good part of the front desk's day.
Dental clinics
Lots of repeated questions about treatments, first visits and financing, and appointments requested after hours.
Aesthetic medicine
Patients who message on Instagram or WhatsApp at any hour to ask about a treatment before booking.
Physiotherapy and multi-room centers
Schedules that change daily, cancellations and gaps nobody fills in time.
Groups with several locations
Each center with its own hours and services, and a front desk that has to know where each patient belongs.
What it does
Appointments
Finds out which treatment the patient wants and when suits them, and creates the appointment request or books it in the calendar if the clinic allows it.
WhatsApp and website
Answers on the clinic's WhatsApp and in the website chat with the same knowledge and the same tone.
Calls
It can also run on the phone, to pick up the calls the front desk misses and leave the message or appointment in the system.
Frequently asked questions
Hours, address, treatments, first visit, payment options. Only what is in the clinic's information, with no made-up prices.
CRM connection
Every conversation enters the CRM with name, contact, reason and channel. We do it with GoHighLevel or with the system you already use, if it has an API.
Hand-off to a person
If the patient asks something clinical, complains or asks to speak to someone, the conversation goes to the team with all the context.
What a conversation looks like
A patient writes on a Sunday night asking whether you offer clear aligners and what the first visit costs. The receptionist answers with what the clinic has published, says the first visit is free when it is, and offers to request an appointment. If the patient agrees, it takes their name, phone number and preferred time, and first thing on Monday the team has it in the CRM.
If in that same conversation the patient says a tooth has hurt for three days, the receptionist gives no opinion. It says a professional needs to assess that, gives the clinic's phone number and flags the conversation as a priority.
What it does not do
It does not diagnose, recommend treatments or interpret symptoms, tests or medication. It does not give prices the clinic has not published or confirm availability it cannot check. In an emergency, it tells the patient to call the clinic or the emergency number.
Those limits are written down before we build anything and tested with trick questions before patients see it. It is the part of the project that takes us the longest, and the one that matters most.
EU AI Act: patients must know they are talking to an AI
Since August 2, 2026, Article 50 of the EU AI Act requires informing people that they are interacting with an AI system unless it is obvious. The receptionist says so in its first message and again if asked. On the phone, it says so when it picks up.
More detail in what the AI Act already requires.
GDPR and health data
Health data is a special category under the GDPR (Article 9) and needs more care than a sales form. That is why we design the receptionist to collect the minimum: name, contact, reason for the appointment and channel. We do not ask for medical history or symptoms, and if the patient shares them, they are not kept beyond what is needed to pass the message on.
Before launch we review with the clinic, and with its data protection adviser if it has one, what data is processed, which model processes it, where it is stored and which data processing agreement is needed. On a health platform where we work as fractional CTO, that review went as far as a data protection impact assessment, which you can see in the Vértika case.
Try it with your clinic's information
The best way to decide is to talk to it. You can try our demo with a fictional dental clinic right now, in Spanish: book an appointment, ask for a price or describe an emergency. To see it with your own clinic, we prepare a private demo with the public information on your website: your treatments, your hours and how to book. You test it with your own questions, including the hard ones, and tell us what you would change. To request it, write to us with the clinic's website.
If you first want to see an AI working within the limits we describe, the assistant on this website works the same way: it says it is an AI, answers only what it knows and offers to pass you to a person. And if you want to know what ChatGPT or Gemini say when someone asks about clinics like yours, request the free AI visibility audit. In our pilot study on clinics' AI visibility, 46 of 78 clinics in Madrid, Valencia and Málaga did not appear in a single answer.
How we work, phase by phase
Clinic information
1 week
We gather treatments, hours, locations, how appointments are booked and the questions the front desk gets most.
Limits and testing
1 to 2 weeks
We write down what it can and cannot say and test it with real questions and trick questions.
Channels and CRM
1 to 2 weeks
We connect it to WhatsApp, the website, the phone if you want, and the clinic's CRM or calendar.
Supervised launch
Ongoing
For the first weeks we review conversations with the team and adjust whatever is needed.
What you get
- The receptionist running on the channels you choose.
- A knowledge base with the clinic's information that the team can update.
- The limits in writing: what it answers, what it hands off and how it acts in an emergency.
- The Article 50 AI notice on every channel.
- The data protection review: what is collected, where it is stored and with which provider.
- An estimate of the monthly running cost with your volume of messages and calls.
How it is priced
Fixed quote after a free diagnosis. There are two costs. Implementation is paid once and depends on the channels, the locations and the integration with your CRM or calendar. Running costs are monthly and depend on the volume of conversations and calls and on the AI model used; we estimate them with your numbers before we start.
Other price references are on our pricing page.
What it does
for you.
- Patients get an answer at night and at weekends too.
- The front desk stops answering the same thing ten times a day.
- Every appointment request reaches the CRM with its source and reason.
- It meets the AI Act notice requirement from the first message.
Further reading
AI agents for companies: what they really automate and what they don't
The difference between a chatbot, an agent and an automation.
Read the articleThe EU AI Act: what already applies to you
The notice that you are talking to an AI, and what was pushed to 2027.
Read the articleImplementing GoHighLevel in Spain
The CRM where the receptionist's appointment requests usually end up.
Read the article
Every project is run by the same team, with the technical side led by Carlos Dodero, lecturer in Artificial Intelligence at Universidad Alfonso X el Sabio. Meet the team or, if you want to start with something concrete, request the free AI visibility audit: we show you what ChatGPT, Gemini and Perplexity say about your business.
Frequently asked questions
An assistant that serves patients on WhatsApp, on the website or by phone the way the front desk would: it answers about treatments, hours and how to book, takes the details of anyone who wants an appointment and logs them in the CRM. It only uses the information the clinic gives it and hands off to the team whatever it should not handle.
No. We design it not to diagnose, recommend treatments or interpret symptoms. If a patient describes a health problem, it says a professional needs to assess it and passes the conversation to the team. In an emergency, it tells them to call the clinic or the emergency number.
Yes. Since August 2, 2026, Article 50 of the EU AI Act requires informing people that they are interacting with an AI system. Our receptionist says so in its first message, or when it picks up on the phone.
Health data is a special category under the GDPR. The receptionist collects the minimum needed to handle the appointment and does not ask for medical history or symptoms. Before launch we review with the clinic what data is processed, where it is stored, which AI provider is used and which data processing agreement is needed.
If your system has an API or can be connected through tools like n8n or Make, yes. If not, the receptionist logs the request in a CRM such as GoHighLevel and the team confirms the appointment. We check this in the diagnosis before quoting.
Fixed quote after a free diagnosis. Implementation depends on channels, locations and integration, and the monthly running cost on the volume of conversations. We estimate both with your numbers before we start.
Yes. There is an open demo with a fictional clinic at elevatectechlearn.com/demo/clinica-dental-ejemplo and we also prepare a private demo with the public information on your website so you can test it with your own questions before deciding anything.
Ready to work together?
Schedule a meeting or tell us about your project. We respond quickly, no strings attached.