How a NZ Plumbing Company Recovered $90,000 in Uninvoiced Work with an AI Operating System

This client is a residential and commercial plumbing company in the trade services industry. They operate across Auckland, Wellington, and Christchurch.

The work was there. The jobs were getting done. But the business side of running a trade company at that scale, the admin, the inboxes, the invoicing, the follow-up, was sitting largely unmanaged. Not because nobody cared about it. Because there were not enough hours in the day to stay on top of it all while also running the actual operation.

The problem

When you run a trade business across three cities, the administrative surface area is bigger than most people expect.

Every job generates communication. Quotes go out, clients respond, suppliers send updates, staff ask questions. Across 11 separate inboxes spread throughout the business, that communication was piling up daily with no real system for handling it. Important messages were getting buried next to spam. Follow-ups that needed to happen were not happening because nobody had a clear view of what was sitting there and what needed a response. The team was spending time managing inbox chaos instead of moving jobs forward.

The invoicing situation was a separate problem with a direct dollar figure attached to it.

Completed jobs were sitting in Fergus, their job management system, without invoices raised against them. The work had been done. The revenue had been earned. But because raising invoices and following up on outstanding jobs required someone to manually go through the system, check job status, draft the invoice, and chase the client, it was constantly getting pushed to later. The result was approximately $90,000 in completed work that had never been billed.

For a trade business, that is not a cash flow problem caused by the market or slow clients. That is money the business had already earned sitting uncollected because the follow-up process did not exist.

Beyond the inboxes and invoicing, there was a broader visibility problem. With teams in three locations, the owners had no clean way to see what was happening across the operation without chasing it down. Jobs moved through Fergus. Messages came in through email. Information lived in different places and required manual effort to pull together into any kind of coherent picture of how the business was actually running on a given day.

What we built

We built an AI Operating System that connected the tools they were already using and put automated workflows on top of them. Fergus, their 11 inboxes, and their invoicing software all became part of a single operating layer that runs in the background without anyone having to manage it.

Inbox management

All 11 inboxes are now managed by the AI Operating System. Every incoming message gets read, categorised, and actioned appropriately without a human having to open it first.

Spam gets filtered out automatically. Urgent messages get flagged immediately so the right person sees them. Routine enquiries get drafted responses that just need a quick review before sending. Client follow-ups that have gone quiet get picked back up and chased without anyone having to remember they were waiting. What used to be a daily flood of unmanaged communication across the business is now a clean, prioritised queue where everything that needs attention is visible and everything that does not is already handled.

Invoicing and job follow-up

The AI Operating System connects to Fergus and their invoicing software and gives the founder a live view of what has been completed but not yet billed. That visibility alone was a significant unlock. Before, nobody had a clear picture of how much completed work was sitting there without an invoice against it. The number turned out to be $90,000.

With that surfaced, the founder could work through the backlog directly via Slack. Ask the bot which jobs are uninvoiced, have it pull the job details, draft the invoice, and send it, all without leaving the conversation. What previously required someone to manually cross-reference Fergus against the invoicing system job by job became a conversation they could work through in an hour.

The same applies going forward. Instead of completed jobs silently accumulating without invoices, the founder has a clear view of what needs to be billed and can action it on demand, with the AI doing the drafting and sending.

Operational visibility

With Fergus connected into the AI Operating System, the owners no longer need to chase status updates across three cities. Daily summaries go out automatically with a clear picture of what is active, what is completed, what is overdue, and what needs attention. The information that used to require three separate conversations with three separate site leads now arrives without asking.

The result

The business did not get bigger. The same team, the same jobs, the same three cities. What changed was how much of the administrative and operational overhead was running itself versus requiring someone's time to push it forward.

The $90,000 recovery was immediate and tangible. But the ongoing value is the visibility that made it possible. The founder now has a clear picture of what has been completed and what has been billed at any point, and can action outstanding invoices directly through Slack in minutes. The revenue the business earns no longer quietly accumulates in a system nobody is checking.

The inbox situation changed the day-to-day experience of everyone involved. Instead of managing communication chaos across 11 inboxes, the team deals with a filtered, prioritised view of what actually needs them. The noise is gone. The important stuff is surfaced. Follow-ups go out without anyone having to write them.

For a trade business at this scale, those are not small things. Admin and invoicing overhead is one of the main reasons trade companies stop growing at a certain size. Not because they run out of work, but because the operational cost of managing more volume gets too high without adding more people. An AI Operating System changes that equation.

The takeaway

Most trade businesses have the same underlying problem this one did. The work is getting done. The jobs are coming in. But the back end of the business, the communication, the invoicing, the follow-up, is running on manual effort and falling through the cracks in proportion to how busy things get.

The $90,000 sitting uninvoiced was not an anomaly. It is what happens when a growing business has no easy way to see what has been done versus what has been billed. The gap does not show up until someone finally looks.

If your business has outstanding invoices that have not been followed up, inboxes that nobody fully owns, or jobs that fall through the cracks between completion and billing, book a strategy call and we will show you what an AI Operating System looks like for your operation.