An AI social media agent for a medical clinic builds the posting calendar for Facebook, Instagram and TikTok, writes the copy, prepares the visuals and keeps track of comments and messages. Posts start from the clinic’s doctors and services. Any message with health details goes to a person, and the doctor approves what is said about treatments.


A clinic’s Facebook page is often run by someone at the front desk, in whatever time is left over. They post when they can, a holiday greeting, a team photo, then nothing for a few weeks. Meanwhile, the inbox fills up with price questions, booking requests and sometimes descriptions of symptoms that nobody answers. The agent takes over the posting rhythm and the order in the inbox.
Builds the monthly posting calendar, with doctor introductions, explanations of investigations, prevention days and clinic news
Writes copy adapted to each platform, longer for Facebook, more visual for Instagram, short and direct for TikTok clips
Prepares visuals in Canva on the clinic’s templates and scripts for clips filmed by doctors
Sorts comments and messages, answers administrative questions and passes on anything to do with health
The conversation with patients stays largely with the clinic. The agent keeps things in order so nothing gets lost, but it doesn’t take the place of the front desk or the doctor.
Tell us what slows you down. We talk, then decide together whether and how we work.

On Monday the agent puts the week’s already approved plan into motion and prepares the materials that need someone from the clinic. If Thursday has a post about the new endocrinologist, it sends the doctor on Monday the script for a short clip and a few questions they can answer in front of the phone camera. If the doctor has no time, the post moves, rather than being replaced with a text about them written without them.
During the week, posts go out at the set times from Meta Business Suite and TikTok. The agent goes through comments and messages several times a day. Questions about hours, address, parking or the price of a consultation get answers from information the clinic has approved. Booking requests reach the front desk or the CRM, with the person’s name and the service requested.
On Friday the agent shows what worked: which posts brought messages or booking requests, which topics sparked questions and which comments needed someone from the clinic to step in. That is where the proposals for the following week come from.

On a clinic’s page, people write things they wouldn’t write anywhere else. They describe symptoms in public comments, ask whether a result is normal and share how they felt after a procedure. Some ask on behalf of a relative, others complain about the wait at the front desk. Here the agent works by fixed rules.
It never responds to the medical content. A public comment with health details gets a short, neutral reply inviting the person to write privately or call the clinic, without repeating what they wrote and without confirming they are a patient of the clinic. The message reaches the front desk or the doctor, flagged as appropriate. If it describes a situation that looks urgent, the person is immediately told to call 112.
Negative comments about a treatment get no reply from the agent. It flags them to the manager right away, with a suggested calm reply that doesn’t discuss the case in public. Medical confidentiality applies on Facebook too.

Meta Business Suite, for posts and messages on Facebook and Instagram, including the shared inbox
TikTok, for short clips scheduled from the clinic’s account
Canva, with the clinic’s brand templates, and CapCut for editing clips filmed by doctors
The clinic’s CRM or booking system, such as GrowClinic, EasyWeek or Zarina CRM, where requests received by message land
Approved photos and clips live in a shared Google Drive folder, separate from any folder holding patient data. The agent has no access to the medical system, only to marketing materials. Posts are scheduled only once the material has all its approvals logged on the calendar.

It doesn’t give medical advice in comments or messages, not even advice that seems trivial
It doesn’t publicly confirm that someone is a patient of the clinic
It doesn’t post images of patients without written consent
It doesn’t get into arguments about treatments, vaccines or other clinics
When a health topic goes viral and people ask for the clinic’s view, the agent prepares a draft and sends it to the doctor. The clinic decides whether to respond and how. Nor does the agent delete critical comments to tidy up the page. It hides only spam, insults and personal data exposed in public, and the clinic sets those rules.

You start with a single page, usually Facebook or Instagram, and with written rules for messages: which questions get a direct answer, what goes to the front desk and what goes to the doctor. The agent receives the approved information on hours, services and prices and prepares the first one-month calendar.
After the first month you see which posts brought questions and bookings and how well message sorting works. Then you add TikTok or doctor clips, at the pace at which doctors have time to film. The message rules are adjusted after the first few weeks, based on the real cases that came up, not on scenarios imagined in advance.
The manager or the marketing person approves the monthly calendar. Posts that explain a condition, an investigation or a treatment are approved by a doctor before they are scheduled. Clips featuring a doctor are approved by that doctor, because it is their face and their name.
Any post with a patient, from a photo of a child to a before and after picture, needs written consent from the patient or the parents and the clinic’s approval. The agent checks that the consent is in the materials folder before scheduling the post.
Replies to criticism, to messages with health details and to any situation that could go public in an unpleasant way stay with people. The agent finds them quickly and puts them in front of the right person, but it doesn’t close them itself. People also decide what happens when the page gets a wave of comments, for example after a local news story about the clinic.
The most important link on the page. From here you go up to the agent we sell.
Sideways, on the same level, so you can move easily between pages.
It answers administrative questions, such as hours, address, the price of a consultation or the documents needed, using only information the clinic has approved. Anything with health details goes to the front desk or the doctor, flagged. Booking requests reach the CRM or the front desk, and someone in the clinic confirms the appointment.
It doesn’t reply on its own. It flags it to the manager right away, with a suggested calm reply that doesn’t discuss the case in public and invites the person to talk directly. The clinic approves the final reply, usually after the doctor involved has seen the situation. The comment isn’t deleted just because it is unpleasant.
It isn’t required. The clinic can post without doctor clips, using explanations, images from the clinic and news. If a doctor wants to appear, the agent prepares the script and the questions so filming stays short, and editing is done in CapCut. The doctor approves the clip before it is published.
Only with the patient’s written consent, without details that identify them if they don’t want that, and with the doctor’s approval. The platforms also have rules for such images, especially in ads. The agent checks that consent exists before scheduling the post and never uses images found elsewhere.
Tell us what slows you down. We talk, then decide together whether and how we work.