Case Study · Beauty & Aesthetics

A touring beauty clinic’s bookings, run by a messaging assistant and a live dashboard.

A travelling beauty clinic runs cosmetic procedures as short, intense camps that move from place to place. Taking every booking by hand over chat couldn’t keep up. JuneAI built an AI front-desk that runs the whole booking conversation, plus a real-time dashboard the team controls from their own server.

24/7
Bookings taken automatically
Multi-site
Camps run from one place
Real-time
Live team dashboard
Self-hosted
Bookings stay on their server

The challenge

The clinic runs as a travelling operation: a camp opens in a location for a few days, fills a tight grid of short appointment slots, then moves on. Every booking arrived as a chat message - which procedure, which day, which time - and a person had to answer each one, check what was still free, hold the slot, collect payment, and confirm. During a busy camp that meant hundreds of overlapping chats and a constant risk of double-booking.

Off-the-shelf booking apps didn’t fit a clinic with no fixed address that moves location every week, prices that vary by procedure, and a customer base reached almost entirely through chat and social media. And handing the whole customer list to a third-party booking platform was something the clinic wanted to avoid.

What we built

A chat booking assistant that runs the entire conversation in the clinic’s own voice - greets the customer, asks which procedure, offers the real open slots for the current location, holds a slot, sends the payment QR, and confirms the appointment once the team approves. It also handles reschedules, cancellations, walk-ins, voice notes, and social-ad enquiries.

A live operations dashboard that gives the team one screen for every camp: real-time booking counters, one-click confirm or reject on pending bookings, full booking tables with location and status tabs, and manual entry for phone or walk-in customers - all updating instantly as new bookings land.

A location & tour manager that opens and closes each location in a couple of clicks - setting the venue, dates, timings and the appointment grid - and keeps an interest list for places with no camp yet, so the clinic can broadcast to a waiting audience the moment a new tour is announced.

Behind it all, a single database is the source of truth for every booking, conversation, location and broadcast - with an audit trail kept on the sensitive actions.

Local-first & data security

The booking database and the team dashboard run on the clinic’s own server, not a booking-SaaS account. The dashboard is reachable only over the clinic’s own private network - it is never exposed to the public internet - so the team manages every camp without the customer list living on a third-party platform.

The clinic owns the system outright: no per-booking fees, no seat-based subscription, and no vendor that can change its terms or hold the appointment history hostage. As the operation adds new locations and procedures, the tool changes with it.

The result

The assistant now takes bookings around the clock, across whichever locations are live, without a person watching every chat - while the team keeps full control through a dashboard that updates in real time. Opening a new location is a two-minute job instead of a manual scramble, and the clinic’s entire booking history stays on infrastructure it owns.

“It’s like having a front-desk receptionist who never sleeps and never double-books. Customers get answered instantly at any hour, and we still approve every booking ourselves from one screen. Best of all, it’s ours - running on our own server, not rented from some app.”

Owner · travelling beauty clinic
Platform: Messaging assistant + self-hosted web dashboard · Stack: Node.js · LLM-driven AI agent · Local database as source of truth · Private-network only
Next case · Construction
Estimating team: 45-minute quotes down to 5
← All case studies

Got a process like this?

If your team is re-typing the same data every day, there’s a tool waiting to be built - and it can run entirely on your own machines.

Start a Project