Replace the carbon-copy work-order book with a phone.

A fire-safety inspection business runs on two things a lot of field-service software gets wrong: recurring compliance cycles that repeat on a fixed schedule regardless of season, and a paper trail that has to hold up if an insurer, a fire marshal, or a client’s own lawyer ever asks for it. We built a system for exactly this kind of business: a small (3-4 person office) fire-protection inspection and servicing contractor running entirely on paper, phone calls, and spreadsheets before this project.

What paper actually cost this business

The technician’s paper form alone had 28 distinct fields: arrival and departure times, break and travel time, hours split across regular, 1.5x, and 2x pay rates for up to two technicians on one job, a safety checklist (was the monitoring agency informed, was the system disabled before work, were keys required and where), work performed, materials used, and a signature. Each completed form produced three carbon copies: customer, office, and technician pay record. The office copy came back to be manually transcribed, originally planned as an OCR-assisted pipeline before the decision was made to skip photograph-and-transcribe entirely and go fully digital instead.

Invoices were built from an Excel template, quotes from a Word template with boilerplate scope-of-work paragraphs assembled by hand for each site. Neither had a designed numbering system. A historical sweep of the client’s own numbering found invoices had reached the 5,500s by late 2025 before restarting in the 2000s for 2026, with quotes running three separate concurrent numbering blocks in the same year.

The accounts-receivable side was worse. Two hand-maintained Excel workbooks held the entire receivables ledger. An independent, adversarial verification (five separate re-derivations of every figure from the raw files) found roughly $1.7 million in nominal outstanding invoices across roughly 490 rows, with invoice numbers reused across genuinely different documents rather than serving as a reliable key at all. A meaningful slice of the book, worth tens of thousands of dollars, was hidden in the spreadsheet while still being counted in every total. A double-digit number of invoices showed up as both paid and outstanding simultaneously. This is not a knock on the client’s staff. It’s what happens to any receivables ledger that lives in a spreadsheet long enough without a system underneath it.

Digital work orders, done once, on the technician’s own phone

The system we built has technicians complete the entire work order on their phone on site: clock in/out with GPS captured when available (never blocking the job if it isn’t), the same safety checklist that used to be on paper, work performed, materials, an optional signature, and a completion photo. No paper book, no photograph-and-retype step, no separate pay-record copy to reconcile against the customer copy later.

A completed, office-reviewed job becomes a one-click invoice: billing information pulled straight from the accepted quote, tax calculated, and an immutable PDF generated and emailed automatically, with delivery status (sent or failed) tracked per invoice rather than assumed. A job only closes once it’s paid in full and that payment has been reconciled against the invoice.

Recurring contracts are the business, so the schedule treats them that way

Monthly and quarterly maintenance and inspection contracts are the backbone of a fire-safety inspection business, not an occasional add-on. We built a recurring-contract scheduling engine that surfaces these jobs on the calendar automatically, rather than depending on office staff to remember a client’s inspection cadence. It shipped only once the client explicitly asked for it in writing, following the same discipline the whole project ran on: build what’s actually needed, not what looks impressive on a feature list.

Roles that keep money away from the people who don’t need to see it

A small office handling intake, scheduling, dispatch, and invoicing all in the same system creates an obvious risk: someone who only needs to book appointments can also see what a client was charged. The system uses three roles with permission boundaries enforced at the structural level, not just hidden in the interface: an admin sees everything, a coordinator handles intake, scheduling, and quotes but is blocked from ever seeing invoices or payments, and a technician sees only their own assigned jobs, losing access entirely once a job is invoiced. In the real deployment, that boundary was checked with direct tests against the system itself, confirming a blocked role actually gets refused access rather than just not seeing a button for it.

A 42,000-file archive, not a filing cabinet

Years of invoices, quotes, reports, and correspondence didn’t disappear when the paper process ended. Roughly 42,000 legacy files were imported and organised into a per-client, per-year document archive, browsable in the client’s own filing language rather than generic system-generated file names, so historical records stay reachable instead of orphaned in an old folder structure nobody opens anymore.

GPS auditing that flags, never blocks

Technician location on a fire-safety job matters for accountability (proving a site visit actually happened as logged), but a system that blocks a legitimate clock-in over a GPS mismatch creates more problems than it solves, especially in a building with poor signal. The approach we build is flag-only: a clock-in or clock-out that looks physically implausible gets surfaced for office review, never auto-rejected. It’s a trust-and-verify posture, not a surveillance one.

Where the honest answer is a lighter tool

If your inspection business runs a handful of clients, simple flat-rate pricing, and no significant legacy paper or spreadsheet backlog to migrate, a general field-service scheduling tool may cover what you need without the cost of a purpose-built system. The case for something custom gets strong specifically at the point where recurring contract cycles, a real compliance paper trail, and a messy legacy ledger all need to work together in one system that speaks your business’s own vocabulary rather than a generic one. If you’re not there yet, say so, and a lighter tool is the right call for now.