An interactive AI version of your clinic
Give prospective patients a way to talk to you before they book — even when you or your team aren’t available. It answers their questions, understands what they’re looking for, and guides them toward booking.
Built around your clinic’s approved information. No diagnosis. No guessing.
Live demo
Compare
A second live avatar from a different provider. Compare the look, voice and feel against the one above.
Try the booking step
In the real product the assistant does this mid-conversation. For the demo, pick a slot and watch it land in the clinic calendar.
The clinic’s live calendar. Book above and it appears here. Open in Google Calendar ↗
The problem
Nine in the evening. Sunday afternoon. Mid-procedure. The enquiry lands, nobody answers, and the patient moves on to the next clinic on their list.
Price. Experience. Location. Recovery. Availability. Am I suitable. Your coordinator answers them dozens of times a week, from memory, with slight variations each time.
And rightly so — they are the difference between an enquiry and a booking. But they go home at six, and enquiries do not.
What it does
Pricing, years in practice, clinic location, recovery windows, what to expect on the day. Delivered verbatim from your own wording.
Name, date of birth, reason for enquiry, preferred date. Structured, complete, and written straight into your system.
Against availability you control and slot rules you define. No judgement calls about urgency — that stays with your team.
Consultation, booking, call or a person — configured around your existing booking and lead-management process.
The mechanism
Most AI assistants generate a reply to anything you ask them. For a clinic, that is the whole risk in one sentence.
Ask a generative assistant whether your surgeon has experience with a particular technique and it will produce a confident yes. Nobody lied. A credential claim just went out under your name.
Clinic Foyer cannot do that. Every response is retrieved word for word from an answer set you have written and signed off. Ask something off-list and it says so, then offers a callback. There is no generative fallback, because the fallback is where the fabrication happens.
Where it works
While a patient is researching a treatment, they can ask questions right there instead of leaving to find you.
Someone reading about Botox or fillers gets the conversation on that exact page.
A patient coming from a post or an ad can start talking immediately.
Each campaign can open its own conversation, tuned to that offer.
Not another chatbot
Patient searches. Finds a wall of text. Maybe.
Patient types. Gets a canned answer in a box.
Patient talks. It responds, asks questions back, and the conversation moves toward booking.
How it works
Start from our aesthetic-practice template and edit, or write your own from scratch. Nothing goes live until you’ve signed it off. Every change is versioned.
One snippet on your site. You define booking availability, which questions escalate to a human, and where captured details are written.
Every conversation, every answer version served, timestamped. If anyone ever asks what your assistant was authorised to say last March, you can show them.
Compliance posture
The constraints above are not a limitation we’re apologising for. They are what keeps this outside medical device regulation and inside your data protection obligations.
Questions
It is designed not to be. Clinic Foyer delivers information you have written, captures enquiry details and books appointments. It does not calculate, interpret or triage patient data to influence a clinical decision — which is the line MHRA draws. We hold that line deliberately, because crossing it would mean UKCA marking and a very different product.
No. It has no capability to. It cannot assess suitability, cannot answer “is this safe for me”, and cannot adjust anything it says to a particular patient’s circumstances. Those questions are recognised and routed to your team.
You do. Enquiry details are written into your systems. We are a processor, not a store — which means there is no database of your patients sitting with us to be requested, corrected, breached or subpoenaed.
It says it doesn’t have an approved answer, offers a callback, and logs the gap. Those logged gaps are useful — they show you exactly which answers to write next, based on what patients actually ask.
At any time, and it is the only way anything changes. Edits are versioned with a record of who approved what and when.
That’s the idea. We build it around an interactive representation of the doctor — appearance, voice and communication style — so the conversation feels like your clinic rather than a generic bot. Patients are always told it’s an AI, so there’s no pretence that they’re speaking to a live clinician.
No. A chatbot asks patients to type into a box. This lets them talk and be talked back to, like a conversation — while still asking questions back to understand what they’re looking for.
No. It handles the repetitive questions and first conversations that happen before a booking, so your team spends its time on patients who need a person. When someone wants a human, it hands over.
Yes. Patient interest doesn’t stop at closing time. The conversation is available whenever someone finds you, and anything it can’t handle is logged for your team to pick up.
Send us your website and we’ll show you how an interactive AI version of your doctor could fit into your patient journey.
No obligation. We’ll tell you honestly if it isn’t a fit.