Flagship product family

Appointment & Capacity Platform

Answer every enquiry, fill every chair, and never lose an after-hours booking.

A booking and customer-capacity system for appointment-driven local businesses. It captures enquiries from every channel in the customer’s language, books against real availability under the owner’s rules, and automatically refills cancellations from a waiting list.

Product concept

The Appointment & Capacity Platform is a complete booking and customer-communication product for businesses whose revenue is capped by the number of appointments they can serve: salons, barbers, driving schools, physiotherapy and dental practices, tattoo studios, beauty and wellness, and small clinics.

It is not a single chatbot or a single reminder feature. It is an operating layer that sits between the customer’s first message and the business calendar. It receives enquiries on every channel, understands them in German, English or Albanian, checks real availability and business policy, books or proposes a slot, secures a deposit when required, and keeps the calendar, the customer record and the owner’s numbers in sync.

The unifying insight is that most appointment businesses do not lose money because they lack customers — they lose it because enquiries arrive when nobody can answer, because cancellations leave chairs empty, and because no one measures either problem. The platform closes both leaks with controlled automation and clear human oversight.

Customer problem

Today a salon owner cuts hair all day. The phone rings while both hands are busy. Roughly half of calls happen outside opening hours, and those callers simply try the next salon. There is no record that the enquiry ever existed.

Messages arrive across WhatsApp, Instagram, a web form and SMS. They are answered inconsistently, often hours later, sometimes in the wrong language for a diaspora customer who would prefer Albanian or German.

When a client cancels the 14:00 slot, that hour is usually lost. Filling it would mean phoning through a mental list of "regulars who wanted Saturday" — which nobody has time to do mid-shift.

No-shows are common because reminders are manual. Deposits are rarely taken because setting them up per booking is awkward.

The economic consequence is invisible but large: unanswered demand, empty chairs, avoidable no-shows, and an owner who cannot say how much revenue was lost last month because none of it was ever recorded.

Before and after

5Today

1. Missed call goes to voicemail
2. Caller hangs up and tries a competitor
3. No record the enquiry existed
4. Empty chair at 14:00 after a cancellation
5. Owner never learns demand was lost

The loss is real but invisible — nothing is measured.

5With the platform

1. Every channel answered 24/7 in the customer’s language
2. Booked against real availability under owner rules
3. Deposits reduce no-shows where policy requires
4. Cancelled slots refilled from a ranked waiting list
5. Owner sees answered demand and recovered revenue

Same staff, same calendar — fewer leaks.

	Today	With the platform
After-hours enquiries answeredillustrative	~0%	every enquiry
Response timeillustrative	next morning	instant
Freed slot after cancellationillustrative	usually lost	offered in minutes

Text alternative (accessible description)

1. Today: Missed call goes to voicemail → Caller hangs up and tries a competitor → No record the enquiry existed → Empty chair at 14:00 after a cancellation → Owner never learns demand was lost
2. With the platform: Every channel answered 24/7 in the customer’s language → Booked against real availability under owner rules → Deposits reduce no-shows where policy requires → Cancelled slots refilled from a ranked waiting list → Owner sees answered demand and recovered revenue

How it fits together

1 / 6

CustomerChannelAIBusiness rulesBusiness systemData storeExternal serviceStaff

Stage 1: Enquiry arrives — A call or message lands on any channel.

Text alternative (accessible description)

1. Enquiry arrives — A call or message lands on any channel.
2. Understand & translate — Speech and text become structured intent in the customer’s language.
3. Check history & availability — Retrieve the customer and the real open slots.
4. Apply policy & book — Deterministic rules and deposits decide the slot.
5. Update systems — Calendar and customer record stay in sync.
6. Confirm & report — Customer, staff and owner metrics are updated.

End-to-end workflow

1 / 10

1. Missed call at 19:40Channel

   Forty minutes after closing, the phone would normally go to voicemail.

2. AI receptionist answersAI

   Greets the caller and detects Albanian.

3. Transcribe the requestAI

   “Men’s cut, Saturday morning.”

4. Match returning customerAI

   Number matches a record; usual barber retrieved.

5. Check availabilityBusiness rules

   Real Saturday slots for a 30-minute service.

6. Apply deposit policyBusiness rules

   Waived for a returning customer in good standing.

7. Provisional bookingBusiness system

   Slot held; confirmation prepared.

8. Customer confirmsCustomer

   Picks 09:30 in Albanian.

9. Schedule reminder & reviewAI

   Friday reminder; Saturday review request queued.

10. Dashboard updatesBusiness system

    Recovered after-hours revenue is now visible.

ChannelAIBusiness rulesBusiness systemCustomer

Step 1: Missed call at 19:40 — Forty minutes after closing, the phone would normally go to voicemail.

Text alternative (accessible description)

1. Missed call at 19:40 — Forty minutes after closing, the phone would normally go to voicemail.
2. AI receptionist answers — Greets the caller and detects Albanian.
3. Transcribe the request — “Men’s cut, Saturday morning.”
4. Match returning customer — Number matches a record; usual barber retrieved.
5. Check availability — Real Saturday slots for a 30-minute service.
6. Apply deposit policy — Waived for a returning customer in good standing.
7. Provisional booking — Slot held; confirmation prepared.
8. Customer confirms — Picks 09:30 in Albanian.
9. Schedule reminder & review — Friday reminder; Saturday review request queued.
10. Dashboard updates — Recovered after-hours revenue is now visible.

Detailed real-world examples

Happy path

After-hours booking that would otherwise have been lost

A two-chair barbershop in Vienna, owner plus one employee. — A missed call at 19:40, forty minutes after closing.

1. The AI receptionist answers on the missed-call fallback and greets the caller. Detecting Albanian, it continues in Albanian.
2. The caller says he wants a men’s cut on Saturday morning. Speech is transcribed and the intent "book · men’s cut · Saturday AM" is extracted.
3. His phone number matches an existing customer record; his usual barber and past services are retrieved.
4. The booking engine checks real Saturday availability for that barber and the 30-minute service duration, and offers 09:30 or 10:15.
5. Because he is a returning customer in good standing, the deposit rule is waived. He picks 09:30.
6. The slot is provisionally held; a confirmation is sent in Albanian with an add-to-calendar link.
7. The owner’s console shows the new booking flagged "auto-booked overnight" for a quick morning glance.
8. A reminder is scheduled for Friday evening, and a review request is queued for Saturday afternoon.

Outcome: A booking that would have gone to a competitor is captured with zero staff effort, in the customer’s language, against correct availability.

Exception path

An ambiguous request is held for human review, not guessed

The same barbershop. — A WhatsApp message: "Etwas für Samstag, vielleicht auch Bart, wie letztes Mal beim Kollegen."

1. The message is understood in German, but the service is ambiguous ("something", "maybe beard", "like last time with your colleague").
2. The system extracts what it can and marks the service and the preferred barber as low-confidence.
3. It does not invent a booking. It creates a draft enquiry, retrieves the customer’s last two visits, and proposes the most likely service.
4. It sends one clarifying question in German: "Gerne! Meinst du Haare + Bart (45 Min) bei Ardit? Dann hätte ich Samstag 11:00 frei."
5. The customer confirms. Only now is the slot booked and the confirmation sent.
6. The interaction is logged so the owner can see how many enquiries needed a clarifying question.

Outcome: Uncertainty is surfaced and resolved with a single question rather than a wrong booking — the customer feels understood and the calendar stays trustworthy.

Different market

Cancellation refilled from a waiting list at a physiotherapy practice

A three-therapist physiotherapy practice in Graz. — A patient cancels a 14:00 slot at 09:00 the same day.

1. The slot is released and immediately marked free in the calendar.
2. The system ranks the waiting list: patients who asked for "this week", who are geographically close, and whose treatment matches the freed therapist.
3. Business rules check eligibility — no double-booking, correct treatment type, and a minimum offer window so nobody is spammed.
4. A timed offer goes to the top three candidates in their preferred language: "A 14:00 slot opened today with Frau Berger — reply YES to take it."
5. The first to accept locks the slot; the other offers close automatically to avoid conflicts.
6. The calendar and patient record update, and the front desk sees the refilled slot without lifting the phone.

Outcome: An hour that was almost certainly lost becomes billable again, with no manual phone-around and no risk of double-booking.

Who is responsible at each step

1 / 9

CustomerAIBusiness rulesSystemsStaff

Step 1 — Customer: Cancels 14:00 slot — Same-day cancellation frees an hour.

Text alternative (accessible description)

1. Customer: Cancels 14:00 slot — Same-day cancellation frees an hour.
2. Systems: Slot marked free — Calendar releases the hour.
3. AI: Rank waiting list — Best-matched patients scored.
4. Business rules: Check eligibility — Treatment, therapist, offer window.
5. AI: Send timed offers — Top three, in their language.
6. Customer: Customer accepts — First YES wins.
7. Business rules: Lock & close others — No double-booking.
8. Systems: Calendar + record updated — Systems in sync.
9. Staff: Front desk sees refill — No phone-around needed.

Product modules

Missed-enquiry capturemvp

A fallback receptionist for missed calls and after-hours contacts across phone, WhatsApp, Instagram, web form and SMS, so no enquiry disappears.

Multilingual voice & message receptionmvp

Understands and replies in German, English and Albanian; transcribes voice notes and calls.

Availability-aware bookingmvp

Books against the real calendar with correct service durations, buffers and staff skills.

Reminders & no-show reductioncore

Automated reminders and optional deposits, tuned per customer segment.

Cancellation & waiting-list refillcore

Detects a freed slot and offers it to ranked waiting-list customers under eligibility rules.

Confirmations & rebookingcore

Handles confirmations, reschedules and recurring appointments.

Review requestsextension

Requests a review at the right moment after a completed visit.

Customer historycore

A lightweight record of past services, preferences and no-show history feeding better decisions.

Owner analyticsextension

Shows answered vs missed demand, refilled slots and recovered revenue — making the invisible loss visible.

Inputs and integrations

• Inbound phone calls (via a forwarding/fallback number)
• WhatsApp, Instagram and Facebook messages
• Web-form and website chat enquiries
• SMS and email
• Voice notes and voicemail recordings
• The business calendar (Google/Outlook or a vertical tool)
• An existing customer list or POS export

Users and buyer

Daily user

Front-desk staff or the owner, who sees the day’s bookings and approves anything the system flags.

Process owner

The owner or manager who sets availability, services, deposit policy and the waiting-list rules.

Economic buyer

The owner — the person who feels the empty chairs and the lost after-hours calls in the till.

Technical administrator

Usually us (managed) or a tech-comfortable staff member; connecting a calendar and a WhatsApp number is the whole setup.

Final decision-maker

The owner. Sales cycles are short because the pain and the buyer are the same person.

AI capabilities

• Speech recognition for calls and voice notes
• Language detection and natural-language understanding across DE/EN/SQ
• Intent extraction (service, staff, timing, new vs returning)
• Translation of outbound messages into the customer’s language
• Ranking waiting-list candidates for a freed slot
• Drafting confirmations, reminders and clarifying questions

Deterministic capabilities

• The calendar and real availability are the source of truth — the AI never overrides them
• Service durations, buffers and staff-skill matching are rules, not guesses
• Deposit and cancellation policy are enforced deterministically
• Offer windows and anti-spam limits on waiting-list refills
• Double-booking prevention and slot locking on acceptance

Object lifecycle

1 / 5

State 1: Requested — Enquiry captured.

Text alternative (accessible description)

1. 1. Requested — Enquiry captured. (→ confirm)
2. 2. Confirmed — Slot held, deposit (if any) taken. (→ remind)
3. 3. Reminded — Timely reminder sent. (→ attend)
4. 4. Attended — Service delivered. (→ request review)
5. 5. Reviewed — Review requested and captured.

Human responsibilities

• The owner sets availability, services, prices, deposit and waiting-list policy.
• Staff approve anything the system flags as low-confidence before it becomes a firm booking.
• A human remains accountable for edge cases, complaints and any deviation from policy.
• Nothing medical is decided; clinical prioritisation always stays with qualified staff.

Economic value

• Recovered after-hours demand that today goes straight to competitors.
• Fewer empty chairs, because cancellations are refilled in minutes instead of being lost.
• Fewer no-shows through reliable reminders and optional deposits.
• Time returned to staff who no longer answer and chase every enquiry by hand.
• For the first time, a number for how much demand was lost — turning an invisible leak into a managed metric.

No market statistics or financial promises are implied. Any figures in the visuals above are illustrative examples, not measured results.

Risks, limitations and failure cases

• A wrong booking erodes trust fast — hence low-confidence enquiries are held for a human, never guessed.
• Over-messaging annoys customers; offer windows and frequency caps are enforced.
• Language nuance matters; Albanian and regional German must read naturally, so templates are reviewed by native speakers.
• Voice transcription is imperfect in noisy conditions; the system asks to confirm rather than assume.
• Calendar integration must be rock-solid — a desync is worse than no automation. Deterministic sync and reconciliation are non-negotiable.

Product evolution

Smallest credible first version

A missed-enquiry receptionist for one channel (WhatsApp or missed calls) that books against the existing calendar in one language, with the owner approving each booking for the first weeks.

Professional product

Multi-channel, multilingual reception with availability-aware booking, reminders, deposits and cancellation refill, plus an owner dashboard.

Optional extensions

• Review requests
• Recurring appointments and packages
• Marketing win-back for lapsed customers
• Deeper POS/payment integration

Long-term platform

A shared capacity layer that other verticals (restaurants, clinics, driving schools) plug into, with the same reception, booking and refill engine behind vertical front-ends.

Shared guidance

Generally applicable method — how to validate, pilot, price and keep humans accountable — lives in the shared playbook so these pages stay specific:

• How to validate an idea
• Design a pilot
• Human approval & accountability
• Pricing hypotheses
• Data protection & privacy