A veterinary clinic · Hong Kong

WhatsApp and the appointment calendar: sort enquiries and draft pet recalls

WhatsApp messages and the appointment calendar separate bookings, changes, emergencies and price questions, with per-pet recall drafts.

168.6 hoursestimated time saved a month
TargetTraceablerules shown with the output
TargetReviewedexceptions reach a person

Representative example — a composite of typical Hong Kong workflows, written from the duties in current job ads. Company names, documents and figures are sample data.

A veterinary clinic operates 2 branches. Enquiries arrive through WhatsApp and must be checked against the appointment calendar. The office needs to distinguish a booking from a reschedule, an emergency or a price question, while retaining a separate vaccination recall date for each pet.

At a glance

The files they already use

  • WhatsApp messages
  • Appointment calendar
AI systembuilt around their rules
  • Separate booking, reschedule, emergency and price questions
  • Keep vaccination recall dates per pet
  • Drafts only: a person taps send

Done automatically

  • Enquiry queue, booking drafts and per-pet recall drafts
  • Exceptions for review

Clinic staff review emergencies and every draft; a person confirms bookings and taps send.

How they used to do it

Reception reads each message to decide what the owner is asking, then opens the calendar to find the related booking. A change of appointment needs the existing entry, while a possible emergency must reach clinic staff without being treated as an ordinary booking.

Vaccination recalls need the pet record, not just the owner's contact. Reception checks the recorded date, writes a reminder and reads it again before sending. Copying a message between pets can attach the wrong recall information.

The difficulties

  • Sort the messages Reception reads the conversation repeatedly to identify the request.
  • Check the calendar A booking or change is copied into the calendar manually.
  • Route emergencies Urgent wording can remain inside the general enquiry conversation.
  • Keep per-pet recalls An owner's contact list does not show which pet's vaccination is due.
  • Review before sending Replies and reminders are written again for each conversation.

Their formula, broken down

No clinical outcome or no-show rate is assumed. This example breaks down the estimated office time for messages, recalls, booking entry and reminder preparation.

Their formulaEstimated saving = existing office time − draft review time; use unrounded time × HK$110
  1. Count message handling

    The estimate covers messages across the working days.

    110/day × 26 days × 3 min143 h

  2. Add recalls, entry and reminders

    Recalls 380 × 4 min ≈ 25.3 h; booking entry 520 × 3 min = 26 h; reminders 520 × 1.5 min = 13 h.

    143 + 25.3 + 26 + 13 hAbout 207.3 h

  3. Retain human review

    Checking drafts and the remaining review still take office time.

    110 × 26 × 45 s ≈ 35.75 h; + 3 h38.75 h

  4. Value the estimated difference

    Displayed time is rounded; money uses the unrounded difference.

    About 168.6 h saved; unrounded time × HK$110≈ HK$18,544

Estimated monthly office-time value≈ HK$18,544

What the system produces

sample.alveek.ai/vet-bookingsSample data

Source records

WhatsApp messagesAppointment calendar
SourceRecord
Message categoriesBooking / reschedule / emergency / price
Appointment calendarCheck existing booking before changes
Vaccination recallRecorded per pet; no invented dates

Enquiry queue, booking drafts and per-pet recall drafts

Review sheet (Excel)Output for approval (PDF)
ItemResult / basis
Booking and reply draftsAwait clinic staff review
Emergency enquiryClinic staff decision
Per-pet recall draftUse recorded date; person taps send
Sending statusDraft only

How it was calculated

Estimated saving = existing office time − draft review time; use unrounded time × HK$110

≈ HK$18,544

Flagged for a person to confirm

  • Missing pet records, unclear appointment changes and possible emergencies go to clinic staff. No vaccination date is generated from an assumption.

In this example: The output retains the source records and the rule behind each item. Clinic staff review emergencies and every draft; a person confirms bookings and taps send.

How we helped

We connect the existing source files to a review sheet, with the company's rules recorded beside the results. Clinic staff review emergencies and every draft; a person confirms bookings and taps send.

  1. Sort the messages WhatsApp enquiries are separated into booking, reschedule, emergency and price-question queues for staff review.
  2. Check the calendar Booking and change drafts are checked against the appointment calendar before a person confirms them.
  3. Route emergencies Possible emergencies are surfaced to clinic staff; the system does not provide a clinical decision.
  4. Keep per-pet recalls Each recall draft uses that pet's recorded vaccination date, with missing records flagged.

What the system handles

Sort the messages

WhatsApp enquiries are separated into booking, reschedule, emergency and price-question queues for staff review.

Check the calendar

Booking and change drafts are checked against the appointment calendar before a person confirms them.

Route emergencies

Possible emergencies are surfaced to clinic staff; the system does not provide a clinical decision.

Keep per-pet recalls

Each recall draft uses that pet's recorded vaccination date, with missing records flagged.

Review before sending

Reply and reminder drafts remain unsent until clinic staff check them and a person taps send.

Before and after

TopicBeforeAfter
Sort the messagesReception reads the conversation repeatedly to identify the request.WhatsApp enquiries are separated into booking, reschedule, emergency and price-question queues for staff review.
Check the calendarA booking or change is copied into the calendar manually.Booking and change drafts are checked against the appointment calendar before a person confirms them.
Route emergenciesUrgent wording can remain inside the general enquiry conversation.Possible emergencies are surfaced to clinic staff; the system does not provide a clinical decision.
Keep per-pet recallsAn owner's contact list does not show which pet's vaccination is due.Each recall draft uses that pet's recorded vaccination date, with missing records flagged.
Review before sendingReplies and reminders are written again for each conversation.Reply and reminder drafts remain unsent until clinic staff check them and a person taps send.

How much time it saves

  • Messages: 110/day × 26 days × 3 min = 143 h; recalls: 380 × 4 min ≈ 25.3 h.
  • Booking entry: 520 × 3 min = 26 h; reminders: 520 × 1.5 min = 13 h; existing total ≈ 207.3 h.
  • Review: 110 × 26 × 45 s ≈ 35.75 h + 3 h = 38.75 h; estimated saving ≈ 168.6 h a month.
  • The estimate values office time at HK$110 an hour (about HK$19,000 a month for a clerk, MPF included); it is not an audited cash return.
  • Unrounded time saved × HK$110 ≈ HK$18,544, about HK$18,500 a month.

These are the working assumptions behind the figure, not an audited result.

How a project like this runs

  1. Read the actual source files. We review the existing inputs with the staff who prepare them and identify where each field comes from.
  2. Record the rules. We turn the agreed rules into visible checks, using completed records to compare the output with the office's own calculation.
  3. Run alongside the existing process. The office keeps its normal process while the system handles the same files. Staff compare the results and resolve differences before switching.
  4. Use and maintain it. Once the results agree, staff use the review sheet and handle its flags. Changes to source formats or company rules are checked before they are applied.

Questions owners usually ask

  • Does the system give veterinary advice?
    No. Possible emergencies and uncertain messages are handed to clinic staff for a decision.
  • Does it send WhatsApp reminders automatically?
    No. It prepares drafts; clinic staff check them and a person taps send.
  • Where do recall dates come from?
    From the recorded vaccination recall date for each pet. Missing dates are flagged for staff, never invented.

The results

  • WhatsApp enquiries are separated into booking, reschedule, emergency and price-question queues for staff review.
  • Booking and change drafts are checked against the appointment calendar before a person confirms them.
  • Possible emergencies are surfaced to clinic staff; the system does not provide a clinical decision.
PythonClaudeExcelWhatsAppEmail

Representative example — a composite of typical Hong Kong workflows, written from the duties in current job ads. Company names, documents and figures are sample data.

← Back to all case studies

Does your company have a formula like this?

Send us one real example — a quotation, an invoice or a month-end statement — and we will show you how the system would work it out. The assessment is free.

Get a free assessment →