From Callout to Invoice by Voice: The AI Mate Field Assistant for Gasfitters and Plumbers

The Problem

Gasfitters and plumbers work with their hands. They are under sinks, on rooftops, in tight ceiling spaces. They are not at a desk. But the paperwork that follows every job — the job sheet, the quote, the time log, the invoice, the gas certificate — has always demanded exactly that: sitting down, typing it all up, at the end of a day when they are already tired. For this Auckland-based plumbing and gasfitting company, that meant technicians finishing jobs on site and then spending another 45 minutes back at the office or at home entering the day's data. Quotes got delayed. Invoices went out late. Details got missed. And the operations manager was still coordinating everything by phone and WhatsApp, chasing updates in real time while trying to juggle the next day's bookings at the same time.

The Solution: Voice-First AI Field Assistant

AI Mate built a voice-first field assistant for this business. The interface is pure speech — the same way you would talk to an AI assistant on your phone, but connected directly to job management, scheduling, quoting, compliance, and invoicing. The gasfitter never has to type, never has to open a form, and never has to sit down and enter data at the end of the day. They just speak. The system handles everything else.

• "I've got a job Tuesday morning, 9 Sunrise Road Henderson, gas hot water cylinder" — job created, customer record matched, certification requirements checked, added to the schedule.
• "I'm heading out now" — departure logged, travel time starts, office notified automatically.
• "I'm on site" — arrival confirmed, job timer starts, customer gets an ETA update.
• "Found a faulty flue, replaced the thermocouple, pressure-tested the line, all good" — job notes captured in full, ready for the gas certificate.
• "Quote is $380 parts and labour" — quote drafted and sent to the customer for approval.
• "Client's confirmed, go ahead" — job confirmed, billing active.
• "Job done" — timer stops. Invoice auto-generated with job notes, parts, time on site, and GST. Sent to the client. Gas certificate pre-filled and ready for sign-off.

MCP-Ready: What That Actually Means

The app is built on the Model Context Protocol (MCP), which means it exposes its scheduling data and actions as tools that any MCP-compatible AI assistant can call directly. In practice this means the operations manager — or the branch manager, or an AI assistant they are already using — can interact with the scheduling system in plain language, without needing to open the app.

• "Who do we have available Thursday morning with a current gas cert?" — answered instantly, directly from the scheduling system
• "Book a gasfitter for the Albany callout Tuesday at 9am" — schedules the job, checks certification, confirms availability, and updates the calendar in one step
• "What jobs are still unassigned this week?" — returns the full list with required skills and urgency flags
• "Will we have enough capacity next week?" — triggers the forecast model and returns a plain-language summary with any gaps highlighted

MCP also means the app plugs cleanly into whatever systems the business already uses. Current integrations connect to the existing job management platform, HR leave calendar, and the NZS 5261 gas certificate generator AI Mate built earlier for the same client. Future integrations — accounting, payroll, supplier ordering — can be added as MCP tools without rebuilding the core app.

Results

• 87% reduction in daily scheduling admin time — from 90 minutes to under 12 minutes per morning
• Zero on-site certification conflicts since launch — all mismatches flagged before the job is assigned
• Same-day callout response time cut from 25 minutes average to under 4 minutes
• Team utilisation up from 71% to 84% of available field hours
• Gas certificate processing time reduced by 90% through integrated pre-fill
• Operations manager: ‘I used to spend my whole morning sorting the schedule. Now I check the AI’s suggestions, approve most of them in one go, and I’m done.’

The app is now live. If your trade or field service business runs into the same problems — certification management, crew scheduling, same-day callouts — this is a deployable solution, not a concept. Talk to AI Mate about what an MCP-ready integration looks like for your existing systems.