Automotive /Customer-service agent live field guide · 9 min

Missed-Call Rescue & Booking Agent

Answers or texts back missed service calls in seconds, captures the vehicle and issue, offers real open slots, and writes the booking into the scheduler — routing anything ambiguous to an advisor instead of confirming on its own.

Open the live demo

theater/demos/automotive_missed-call-rescue-booking-agent.html · sandbox · read-only

Open

FIG. 1

The live demo, running on fabricated data. Open it to step through the full flow — every output is shown for a person to approve before anything happens.

How it would work

Reads the missed call's voicemail, classifies the vehicle, symptom and urgency, drafts an SMS that offers real open slots, and routes everything to an advisor to approve before any reply is sent.

Input 01

The missed call + the diary

The missed-call voicemail (auto-transcribed), the caller's number, and the workshop's live booking diary — open slots, technician assignments, job durations.

Agent 02

Classifies, then drafts

Extracts the vehicle, the described symptom, the deadline and an urgency signal, picks a matching open slot, and drafts an SMS reply with the offered time and any safety advice.

Output 03

A draft reply, held for approval

A drafted SMS and a proposed booking, with the offered slot and est. job value shown, queued for a service advisor to approve, edit or escalate before anything is sent or written into the diary.

Where it works well

It catches the calls you lose during the morning lane — fast, and every one of them.

Best for a service desk that misses a real share of calls at peak — the morning lane, the lunch gap, after hours — and can't reliably call them all back before the customer books elsewhere.
It replies in seconds and re-sorts the overnight queue by urgency, so the team arrives to a triaged list instead of a cold voicemail pile.
The recaptured time goes back to the advisor's bench — the diagnosis conversation, the upsell, the customer who's standing at the counter — not into chasing last night's voicemails.

The expensive part of a missed call isn't the call, it's the silence after it: a caller who hits voicemail at 8:52am during the lane rush rings the workshop down the road and books there before anyone returns the call.

Where it works badly

It is confidently wrong when the caller mumbles the rego, the symptom is vague, or the diary is stale.

Weak on ambiguous voicemails — an unclear vehicle, a symptom that could be three different jobs, a caller who half-states the problem. It should flag those for a human, not guess a booking.
If your diary isn't the live source of truth — slots blocked on paper, double-booked by walk-ins — every offer it drafts is a slot it can't actually honour.
It can't hear a safety problem the caller didn't describe; "grinding brakes" it can flag, a silent fault it can't.

The honest test

If you can't trust the slot your own diary shows as open right now, this tool offers your customers a time you can't keep — faster, not better.

A voicemail that says "the thing's making a noise, give us a buzz" gives it nothing to classify — and a slot it offers against an out-of-date diary is a booking you have to unwind with an annoyed customer.

What it doesn't do — and shouldn't

It drafts the reply and proposes the booking. An advisor sends it. That line is deliberate.

WHAT IT DOES

Shows the vehicle, symptom and urgency it read from the voicemail

Names the slot it matched and why (earliest senior-tech, job duration)

Flags safety signals and disputes for a human — and offers a safety advisory in the draft

WHAT IT WON’T

Send the SMS or confirm the booking on its own

Diagnose the fault or commit to a repair price

Reply to a billing dispute, a complaint, or an unknown caller

A reply that downplays a brake fault, or a booking the workshop can't honour, lands on the workshop — not the tool. Under the Australian Consumer Law a workshop's repairs carry consumer guarantees, and a disputed job can end up at a tribunal like VCAT. The advisor stays on the send button because the consequence — safety, the customer relationship, the standing of the business — stays with them.

What your data has to look like

A live booking diary and voicemail you can actually transcribe — the two things most workshops half-have.

Typical readiness

across orgs we see, before the first job

A live, single-source booking diary

Needs shaping

Voicemail captured as audio it can transcribe

Usual weak point

A caller mobile number to text back

Usually ready

A service menu with typical durations

Usual weak point

A routing map of who handles what

Needs shaping

The real first job

The diary is usually the weak point — slots blocked on a wall planner, double-booked by counter walk-ins, or living in a DMS nobody reconciles. Getting the booking diary to be the single live source of truth is usually the real first job, larger and more valuable than the SMS layer on top: once an open slot genuinely means open, every offer the tool drafts is a slot you can keep.

Right fit if…

You miss a real share of service calls at peak — morning lane, lunch, after hours

Your booking diary is live and digital, and an open slot really means open

Missed-call callbacks are inconsistent today and customers book elsewhere

Most calls are routine bookings — logbook, brakes, tyres — not bespoke diagnostics

Walk away if…

Your diary lives on paper or is routinely double-booked by walk-ins

Missed calls leave no recording the tool can transcribe

Most enquiries are complex diagnostics that need a tech on the phone

You want it to confirm bookings and answer disputes without a human

Open questions

The worried-buyer questions, answered straight

It can draft a wrong reply, which is exactly why nothing is sent on its own. It shows the vehicle, symptom and urgency it read, the slot it matched and why, and flags safety signals like grinding brakes for a human — and the draft it writes advises the caller not to drive far until it’s checked. A service advisor reviews and approves before the SMS goes out or the booking is written. The tool drafts; the advisor stands behind it.

It works from what the caller actually says. A clear voicemail — vehicle, symptom, deadline — it can classify and draft against; a vague one it should flag for a human rather than guess a booking. The bigger dependency is the booking diary: if your open slots aren’t a live, single source of truth, even a perfect transcript produces an offer you can’t keep. Getting the diary clean and live is usually the first piece of work, and it pays off on every booking after.

No. It removes the after-hours and peak-lane callback chase — the cold voicemail pile waiting at 8am — so the advisor spends their time on the diagnosis conversation, the customer at the counter, and the judgement calls it routes to them. Every booking still gets a human’s approval before it’s confirmed. The capacity it frees goes back to the bench, not out the door.

Current to the minute it offers a slot. The tool offers times from your live diary, so if slots are blocked on paper, double-booked by walk-ins, or held in a DMS nobody reconciles, it will confidently offer a time you can’t honour — and you wear the customer’s annoyance when you ring back to move it. The honest test: does an “open” slot in your diary right now genuinely mean open?

A caller’s mobile number and the contents of their voicemail are personal information under the Privacy Act, and a workshop’s customer records carry obligations under the Australian Consumer Law. Any deployment runs against your own systems — your telephony, your booking diary — not a shared pool, and we scope where the data sits and who can see it as part of the build. The demo here runs entirely on fabricated data; Mark Davies and his HiLux are not real.

It doesn’t reply to them. A voicemail that disputes an invoice, threatens a tribunal like VCAT, or comes from a withheld number is flagged as “do not auto-reply” and routed to the service manager with the full transcript attached. The tool’s job is to book straightforward service work fast and hand the human everything that needs a human — a complaint is squarely the latter.

What it takes to build

3–4 weeks · 4 phases

Reused from template~65%

Bespoke to this skin~35%

stack · Claude · Twilio · scheduler API · review UI

What it would cost

Fixed scope, fixed price, fixed dates.

Bite-sized first piece

One contained change, low risk

Pilot build

Most builds land here

Embedded support

Scale on proof

Losing service calls at peak?

The honest place to start is a bite-sized first piece — one contained change, low risk. Tell us where the calls are leaking; we'll play it back, scope it, and show you what's possible.

Book a call How we work