ABM ServiceAI Voice Modules
Book a demo
New — for ABM Service customers

Your customers talk. Your engineers talk.
ABM fills itself in.

Introducing the AI voice modules for ABM Service, powered by AgentDesk. This page follows one service call, end to end: a customer reports an alarm fault by voice, the engineer dictates the fix on site, and the customer simply asks how to pay. All three modules are live today — and along the way you'll see, clearly marked, what AI takes over next.

Live now iPhone & Android app Works in the web portal you use today
Real-time voice AI — the same technology behind today's leading assistants Built into ABM — nothing new to install or learn In the app stores — AgentDesk for iPhone & Android
01 The customer reports it — by voice

Your customers describe the problem. ABM does the rest.

No forms, no phone queue, no waiting for office hours. A customer presses one button, says what's wrong — and a properly classified service call lands in ABM with a reference number and an email confirmation.

  • Spoken in, structured out. The rambling description becomes a clean, professional call summary.
  • Call type and status assigned automatically — matched to your own categories, no picking from dropdowns.
  • Instant confirmation. Reference number on screen, professional email in their inbox.
  • 24/7. Works from the web or the mobile app, any hour — your office team stays uninterrupted.
  • The back office gets the full picture. The AI summary, the complete word-for-word transcript — and one click to listen back to the actual call.
Watch: a call logged by voice
50-second demo — filming now

The call is now in ABM. Today, the back office reads the AI summary, plays back the recording, and books an engineer. On the roadmap: AI does that legwork inside ABM Service — and your team just approves.

AI working the call · back office Coming soon
Call history — two earlier faults on this panel
Similar issues — same fault matched at two other sites
Knowledge bases & manuals — searched
Online sources & documents — checked
Draft resolution ready
Likely loose door contact on zone 12 — refit, reseal, test
Best engineer proposed
Mark O'Brien — alarm-qualified, closest, free tomorrow
Nothing moves without your team — one click, and the brief goes to Mark. Approve ✓
The human stays in the loop. The AI does the digging.
See it now — interactive preview ↗
Calling the customer back Coming soon
Agent · speaking
Your issue
Zone 12 tamper fault — service corridor
Engineer
Mark O'Brien
Visit booked
Tomorrow, 9–11 am
Nobody in the office had to make this call.
See it now — interactive preview ↗
02 The engineer resolves it — on site

Engineers hate typing on phones. So we removed the typing.

To close a job, the resolution and the times have to go in — that hasn't changed. What's changed is how: the engineer opens the app and says them. What they did, when they started and finished, the travel — spoken once, captured completely.

  • One tap, then just talk. Start the voice call from the app and say what happened — no forms, no typing.
  • No detail lost. The full story in the engineer's own words, tidied into clean notes automatically.
  • Times that are finally right. "Started at half nine, finished at quarter past eleven" — captured exactly, travel time too.
  • Straight into ABM. Ready for invoicing, reporting and the customer record — the moment they hang up.
Watch: a resolution, spoken
60-second demo — filming now

Fixed, logged, on the customer's record — resolution and times, ready for invoicing. Soon the agent will ring the customer with the outcome too — but one thing they already do today: pay for the work.

03 Every question answered — even "how do I pay?"

Every "how do I…?" — answered on the spot

A help assistant lives inside the ABM Web Portal. Your team — and your customers — ask in plain language, out loud or by typing, and get an accurate answer drawn from the real documentation, in seconds. The job's done, the customer wants to settle up? They just ask — and the assistant walks them to the Make Payment button.

  • Voice or text — their choice. Have a spoken conversation, or type a quick question.
  • Grounded in your actual documentation — answers cite the guide they came from. No made-up replies.
  • Knows when to hand over. If it can't answer, it raises a support ticket with the full context — and emails a summary.
  • Role-aware. Office staff get admin answers; your customers get customer answers — right down to "how do I pay for that call-out?"
  • Engineers use it on site, too. Load a panel's manuals, and the engineer talks through error codes and faults with a voice agent — straight from the app in their pocket.
Watch: "how do I pay?" — answered
40-second demo — filming now

What this changes for you

Voice isn't a gimmick here. It fixes the oldest problems in field service: incomplete notes, wrong times, and calls that never get logged.

Complete job history

No more blank or one-word resolution fields. Every job carries the full story of what was found and fixed.

Accurate times, every job

Start, finish and travel captured correctly at last — the numbers your invoicing and reporting depend on.

Calls logged 24/7

Customers report problems the moment they happen — correctly classified, confirmed by email, no phone queue.

A happier team

Engineers talk instead of typing on a phone keyboard. The office answers fewer "how do I" calls. Everyone wins.

And this is just the first release

The three modules above are live today. Here's where it goes next — because every voice call already captures the description, the full transcript and the recording, the AI has everything it needs to start closing the loop:

Next

Show the fault — photos & video in the call

While describing the problem by voice, the customer snaps photos of the error or records a short video — each one landing in the call timeline at the exact moment it was taken. AI reads them together with the words being spoken and folds what it sees into the problem description. Something a phone line can never offer.

See it now — interactive preview ↗
Then

AI-drafted resolutions in the back office

When a call comes in, an AI assistant reads the problem, checks the customer's call history and similar past issues, searches your knowledge bases — and proposes a resolution. Your support team validates instead of researching from scratch.

See it now — interactive preview ↗
Then

The right engineer, proposed automatically

Every call already arrives classified by type. Next, the system checks the ABM scheduler and proposes who should take the job and by when — weighing each engineer's skills, past jobs, location and availability. Your planner approves instead of assembling.

See it now — interactive preview ↗
Then

The agent calls your customer back

Resolution ready? AgentDesk rings the customer with the full context — what the problem was, how it's being resolved, and which engineer is coming on site — with every detail available if they ask. Plus a resolution email with an interactive link.

See it now — interactive preview ↗

The direction is simple: AI does the gathering, drafting and follow-up — your people validate and decide. Staying with ABM means your system keeps getting smarter, without changing how your team works.

See it live, on your kind of jobs

Twenty minutes, real product, no slides. We'll log a call by voice, resolve it by voice, and answer questions from the documentation — and you can try it yourself.

Book a demo
Or simply reply to the email that brought you here — paul@dpsystems.ie