Skip to content
prom.work
/
Back to all work
Representative scopeAircon & home services

Turning a LINE group and a whiteboard into routes, records, and repeat work

The scope I quote a small service crew: jobs dispatched to each technician's LINE, a service record per machine, and the six-month re-clean reminder that brings the customer back without anyone calling.

FieldAircon & home services
Service linesBooking apps, Custom software & internal tools
TimelineScoped to map, build, and hand off inside a few focused weeks
ProvenanceRepresentative scope
Day routesDDispatch

Aircon cleaning · 2 units

Lat Phrao 71

2 units

Ton

  • 400฿

Boy

The problem

Jobs come in as addresses pasted into a LINE group and the schedule lives on a whiteboard. Nobody remembers which compressor we cleaned last year, and customers only come back if they happen to remember us.

Who it is for

Small crews that work in homes and shops: aircon cleaning and repair, pest control, deep cleaning, appliance service. Built for the four-to-ten person outfit where the owner dispatches, quotes, and chases payment between their own jobs, and where marketplaces bring work but keep the customer.

Scope of work

The build keeps the crew inside LINE, where they already live, and puts a system underneath it.

  • A job board the owner runs from a phone: new job in, assigned to a technician's day route
  • Job cards pushed to the technician's own LINE: address, map link, machine history, photos
  • Customer notifications in LINE: booking confirmed, technician on the way
  • A service record per machine: serial, BTU, last cleaned, before-and-after photos
  • PromptPay collection on completion, with a digital job slip the customer keeps
  • Automatic re-clean reminders on the six-month cycle, with one-tap rebooking

What you own at hand-off

  • The full codebase in the crew's own repository
  • Hosting, domain, and LINE wiring set up in the business's name
  • A plain-language runbook: add a technician, reschedule a day, refund a job
  • A recorded handover call the owner can rewatch instead of re-asking

Package tiers

Essential

For
One crew, one calendar: dispatch and job cards working end to end.

Includes

  • Job board with day routes
  • Job cards into technicians' LINE
  • Customer booking confirmations

Studio

For
Crews that want the paper gone: machine records, photos, and digital job slips.

Includes

  • Everything in Essential
  • Per-machine service records with photos
  • PromptPay collection and digital job slips

Signature

For
Crews ready to grow on repeat work: the reminder cycle becomes the sales team.

Includes

  • Everything in Studio
  • Six-month re-clean reminders with one-tap rebooking
  • Owner dashboard: jobs, routes, and repeat rate

Timeline & phases

  1. Map

    A short first week

    I follow one job from the first LINE message to the cash, and write down every detail that lives only in someone's head.

    A one-page scope: the dispatch flow, the record per machine, and what stays in the group chat on purpose.

  2. Shape

    Inside the same fortnight

    I design the job card for a technician on a motorbike and the board for an owner dispatching with one thumb.

    Clickable screens the owner approves before any backend is built.

  3. Make

    The build weeks

    I build dispatch, the LINE job card, and the machine record as one tested path, then add payment and the job slip.

    A real job dispatched to a real technician's LINE and closed with a real PromptPay payment.

  4. Prove

    A focused week

    The crew runs a real week on the system next to the whiteboard, and we keep score of what the whiteboard still wins.

    A short list of fixes from the field, all closed before the whiteboard comes down.

  5. Hand-off

    The final days

    I move every account into the business's name and walk the owner through a full day, dispatch to job slip, without me.

    The keys: code, hosting, accounts, and a runbook in plain Thai and English.

How pricing works

  • Quoted per project after a short scope call. You see the full number before any money moves.
  • Payment is phased against the build: a deposit to start, a milestone at the first dispatched-and-paid job, the balance at hand-off.
  • Fixed scope, fixed price, and no commission per job. The customer list this builds belongs to the crew, not to a platform and not to me.

Assumptions

  • Technicians have LINE on their own phones; nobody installs a new app in the field.
  • The business has, or will open, a PromptPay-capable account in its own name.
  • One owner-dispatcher makes the calls during the build.

Deliberately out of scope

  • No GPS tracking of technicians; the job card says on the way because the technician tapped it, not because the system spies.
  • No inventory or spare-parts module in the first scope; jobs and records first, stock later if the work proves it.
  • No payroll; the owner pays the crew however they already do.

The one decision that mattered

The reminder is the business model, not a notification

A dispatch board saves hours, but hours are not why this build pays for itself. Aircon work runs on a known cycle: a machine cleaned today wants cleaning again in about six months. The decision that matters is recording every machine well enough that the six-month reminder can fire with the right address, the right unit, and a one-tap rebook. That single loop converts a one-time job into a customer the crew never has to win twice.

I would build the machine record before the dispatch board. A job card with history attached is what makes the crew look like professionals at the door, and it is also the only thing the six-month reminder can stand on.

Dispatch itself stays almost rude in its simplicity: the owner drags a job onto a name, the technician's LINE pings, done. Field crews abandon anything slower than the group chat they already have, so the system has to beat the group chat at its own speed.

The reminder loop ships last and dark: it runs silently against real records for a cycle before the first customer ever sees one, because a reminder about the wrong machine would cost more trust than it earns.

Outcome & what it proves

The result

Built so that the whiteboard comes down, every machine has a history with photos behind it, and the calendar starts filling itself when the six-month reminders fire. The crew keeps the customer relationship that marketplaces would otherwise own.

What it proves

That for service crews, software earns its keep on the second visit, not the first. Record the machine, respect the cycle, and repeat work stops depending on the customer's memory.

Aircon & home services

Have a problem that looks like this?

One message in plain language is enough to start. I map the work, scope it fixed, and you see the full number before any money moves. I reply within one business day.