// case study — retail
A salon that runs on
one system, not six.
Bookings in a diary, payments in a card terminal, packages on a spreadsheet, reminders sent by hand. Kapil Hair Studio replaced that with a single system — including an AI voice assistant that answers the phone while the team is with clients.
// the problem
The busiest hour is the worst time to answer the phone.
A premium salon has a scheduling problem that looks simple and is not. Services have different durations. Some need a specific stylist, others do not. Some need a gap in the middle — colour has to develop — during which the chair is occupied but the stylist is free. Get that wrong and you either double-book a person or leave money sitting in an empty slot.
On top of that sits everything else a salon sells: courses of treatments, gift cards, loyalty, deposits and no-show protection. And the phone rings hardest precisely when every stylist is mid-appointment and nobody can answer it.
// what we built
Scheduling first, because everything else depends on it.
The core is a resource-timeline calendar: stylists as resources across the day, with availability rules, service durations and staggered steps handled properly rather than approximated. Customers book online against real availability, and the same view is what the team works from at the desk.
Around it sits the commercial machinery: service packages with sessions drawn down as they are used, gift cards, a loyalty scheme, stored payment methods for deposits and no-show protection, and card and Apple Pay checkout. Every completed appointment produces a proper PDF invoice, generated server-side so the document is identical whoever opens it.
Transactional email — confirmations, reminders, receipts, rebooking prompts — runs through a dedicated email service with delivery events tracked, so “the client says they never got it” is a question with an answer. There is also a rebooking flow, a recommendations surface, an analytics view for the owner, and a careers section, because a salon group is also always hiring.
// the interesting part
An assistant that answers the phone.
The salon's hardest operational hour is the one where every stylist is with a client and the phone will not stop. Missed calls are missed bookings, and a callback an hour later often finds the caller has booked elsewhere.
So the system includes an AI voice assistant — “Kylie” — that answers calls, explains services and pricing, checks genuine availability, and takes bookings, in a register that matches how a premium salon actually speaks to its clients. It is grounded in a maintained knowledge base of the salon's services rather than allowed to improvise, which matters when the question is about a treatment's suitability or price.
The design rule is the same one we apply to every AI project: the assistant handles the mechanical part — availability, service information, straightforward bookings — and hands to a human the moment a conversation needs judgement. A voice agent that confidently invents a price is worse than a voicemail.
// notes
Built on current versions, deliberately.
The system runs on Next.js 16 and React 19 with Supabase behind it. Staying close to current releases is a maintenance decision rather than a fashionable one: upgrades taken in small steps are routine, whereas a stack left three major versions behind turns every security patch into a project.
The site and system are live at kapilhairstudio.co.uk. As with all our work, commercial performance figures belong to the client; we can arrange a reference conversation rather than publish numbers of our own.
// next step
Tell us what is actually broken.
A short message is enough to start. The first scoping call is free, and you will be talking to the person who would write the code.