AI Customer Service on WhatsApp for Clinics: From Ad to Booked Appointment
The patient who reaches out through the ad or the website gets a reply in seconds, at any hour. The AI asks only what's needed to book (specialty, insurance or self-pay, and shift) and hands off to the front desk, with a deadline to reply. Diagnosis, prescriptions, and treatment guidance always stay with the doctor.
SofiaHello, Renata! This is Sofia, from the Horizon Clinic front desk. Would you like to book an appointment or ask a question?08:12
I'd like to book a routine checkup.08:13
SofiaOf course! Will it be through insurance or self-pay? And do you prefer morning or afternoon?08:13
Through insurance, in the morning.08:14
Juliana, from the front desk, took over
Where the patient slips away today
Messages wait until the front desk has a free minute
01
The front desk handles walk-ins and the phone; WhatsApp waits, and the patient books wherever someone replied first.
02
Anyone who writes at night doesn't get a reply until the next day, by which point they've given up.
03
No-shows and empty slots don't come back: without a reminder, the calendar loses appointments that were already booked.
The patient's journey
A straight line from ad to appointment
Today, messages wait until the front desk has a free minute. Here, the AI replies right away, asks only what's needed, and logs every step on the pipeline card.
AdForm or click to WhatsApp
AI repliesIn seconds, at any hour
Ready to bookSpecialty, insurance, and shift
With the front desk30 min to offer a time slot
Appointment scheduledIn the connected calendar
ConfirmedReminder 24 h before, without naming the specialty
AttendedSyncs back to your system
Done in the dashboardSyncs back to your system
Built for clinics
A front desk with time for the people in the clinic
Front desk notified, with a deadline
Once a patient is ready to book, they reach the front desk with the specialty, health insurance, and preferred shift already noted. The front desk has 30 minutes to offer a time slot, and the dashboard alerts you when the deadline gets close.
JPMACS21:40Juliana has 21 minutes to offer Renata a time slot.
The pipeline the clinic uses
New contact, ready to book, with front desk, appointment booked, confirmed, and attended. The booked appointment requires the specialty, insurance or self-pay, and the date.
New
Schedule
Front desk
Scheduled
Confirmed
Attended
Health safeguards already on
The AI doesn't diagnose, recommend medication, or comment on test results. Specialty and reason for the visit are masked on screen and left out of alerts to other systems and out of exports.
No diagnosis or prescriptions in the chat
Specialty and reason treated as sensitive data
Child appointments: the conversation is with the parent or guardian
Emergency: advises calling 192 and alerts the front desk
From Spend to Completed Appointment
Upload your ad spend spreadsheet and see, by campaign, ad set, and ad, what each contact, each booked appointment, and each completed appointment cost.
SpendR$ 8,800
Contacts503
To book308
Scheduled183
Completed128
Confirmation reminder that doesn't expose the patient
Twenty-four hours before, the AI asks for confirmation on WhatsApp without mentioning the specialty or the reason, because the notification shows on the phone's lock screen. Patients who confirm move to Confirmed; those who can't make it reschedule with the front desk.
Appointment scheduledReminder 24 h beforeConfirmedCan't make it? The front desk reschedules
Examples use fictional companies, people and numbers.
Ready on day one
During onboarding, the specialist sets up your account for the “Medical clinics and practices” industry, and it starts out ready to go
This is what your account gets on first login, taken from the same template that builds the dashboard. You can change everything later, from stage names to questions.
The pipeline
New contact
Ready to book
With the front desk
Appointment scheduled
Confirmed
Attended
The qualification questions
Specialty“Which specialty would you like to book?”
Who the appointment is for“Is the appointment for you or for someone else?”
Insurance or self-pay“Will it be through health insurance or self-pay?”
First visit or follow-up“Is this your first appointment with us?”
Day and time“Do you prefer mornings or afternoons?”
By the numbers
Pipeline stages
6
Deal fields
7
Qualification items
5
Re-engagement messages
3
Industry considerations: Health data is sensitive: the AI asks only what's needed to book. No diagnosis, no treatment advice, and no promised results. This summary guides the setup and is not legal advice.
In one afternoon, without replacing your clinic's system
Connect the clinic's WhatsApp
The official account, with approved templates. The patient gets the first reply within seconds.
Choose the Clinic template
In Pipeline → New board, the board comes with the stages, the fields (specialty and reason flagged as sensitive), and the urgency rule.
Build the front desk's knowledge base
Specialties, accepted insurance plans, self-pay appointment fee, and follow-up visit policy. The AI only answers what is in the knowledge base.
Does the AI give medical advice?
No. It doesn't give diagnoses, recommend or swap medications, or comment on test results. If the patient mentions how they feel, the AI uses it only to find the right appointment slot, without asking for details.
What if it's an emergency?
The AI tells the person to call 192 (Brazil's emergency medical line) or go to the emergency room, and alerts the front desk right away. If the person mentions hurting themselves, the AI responds with empathy, points them to the CVV crisis line at 188, and alerts the front desk.
What about LGPD?
Specialty and reason for the visit are health data: they're masked on screen and left out of alerts to other systems and out of exports. Child appointments are discussed with the parent or guardian, and the AI asks only for the age.
Can I promote the clinic in campaigns?
Yes, as long as the message includes the facility's name and the name and CRM number of the medical technical director, as Resolution CFM 2.336/2023 requires, and doesn't promise treatment results. That reminder comes built into the industry's safeguards, inside your account.