Questions,
answered.

We bridge the gap between your people and the technology. We start with what is actually slowing your team down, not with AI, then build the systems that cut the wasted time and open up new revenue.

The answer looks different in every business, because it starts from how yours already runs.

No. ChatGPT is one tool out of many. Think of a builder's toolbox: ChatGPT is the hammer. Useful, but you would not build a house with only a hammer.

Most New Zealand businesses that say they are using AI mean someone has ChatGPT open in a browser tab. That is a fraction of what is available. The things we build, agents that know your business and automations that move work between your systems without anyone retyping it, are a different category of tool entirely.

Think of it as a digital person who has read everything your business has ever written down.

A person builds up a brain over years of schooling and experience. An AI agent starts with an empty one. We fill it with your documents: handbooks, policies, procedures, price lists, job notes. Then we give it a personality that fits your team. We built one for a trades business that answers like a tradie, short and blunt, because that is what the crew would actually read.

It does not sit on the open internet. It sits in a closed environment only your people can reach.

Wherever your team already works. We deploy agents into Microsoft Teams, WhatsApp, your website, or a private web app, whichever place your people are already in every day.

That is deliberate. A tool your team has to remember to open is a tool your team forgets to open.

Both, but the building is the point. We are a hands-on implementation partner, not an advisor who hands over a roadmap and leaves.

We are also not looking for one-off jobs. The technology gets more capable every few months, so the useful relationship is one where we keep improving what is there as you grow, not one where we drop a solution and walk away.

SMEs, across a wider spread than people expect: electrical, plumbing, accounting, HR, kitchen manufacturing, travel, aged care, and a charity.

Tradies on one end, accountants on the other, and most things in between. If you have admin piling up, work being done twice, or opportunities going missing, you are our people.

It depends on the build, and we map the value before we ask you to spend anything.

Projects run from smaller builds through to full system overhauls. Price follows complexity and the return, so a single automation and a whole operating system are not the same conversation. You will know what it is worth before you commit.

There is a monthly fee as well as the build. It is closer to a website hosting arrangement than a software licence.

The monthly covers three things: the running costs of the AI itself, which we pay on your behalf; ongoing optimisation as your documents and processes change; and the upgrade work each time a new model is released.

What you would pay depends on how heavily it gets used, which is why we map that out before quoting. Someone using an agent as part of their daily work generates very different usage from someone occasionally looking something up.

Because it is not finished, in the way software normally is.

New AI models are released every few months, and each one is faster and more capable than the last. They also behave differently. A new model often needs the way your information is organised to be reworked before it performs properly. That work is on us, and it is a large part of what the monthly covers.

The upside is that what you bought keeps getting better without you buying anything again.

Sometimes, and it is worth asking. There are regional business and innovation funds that support this kind of work. We will look into what applies to you as part of scoping, rather than leaving you to find it.

Not necessarily, and the case for it is often stronger, not weaker.

Organisations running on volunteers have the least slack for admin and the most to gain from removing it. The real question is usually whether an added operating cost can be sustained year after year, and that is a conversation worth having openly before anything is built.

A retainer buys a set number of hours each month for continuous improvement work: automations, opportunity-hunting, training, troubleshooting, and advice.

Each month starts with agreeing priorities and ends with a summary of what was done, hours used, and estimated time saved. Anything large enough to eat the month gets scoped as a separate project instead, so the ongoing momentum does not stall. Hours do not roll over.

Yes. What we build is closed, private and shaped around your systems, so your information is not floating around a free AI tool.

Most of our systems run on a machine we own, here in New Zealand. Our privacy statement sets out exactly what we hold, where it sits, and who else can see it. Nobody should have to guess.

No. What we are using is their processing power, not their consumer apps.

The distinction matters. When you type into the free ChatGPT app, you are using a product that learns from how you use it. When your agent sends a question through the same underlying model on a business connection, the model does the thinking and hands the answer back. It is not keeping your data or training on it. Your information stays inside your environment.

If your business runs on Microsoft, Copilot is the most locked-down option of all, because it was built for businesses from the start.

Yes, mildly, and telling people to stop will not work.

The free tools learn from what you feed them. Client names, pricing, margins and personal details should not go in. But staff use these tools because they genuinely help, so a ban just pushes it out of sight.

Two things actually fix it: a short, plain AI policy so people know where the line is, and a private alternative good enough that nobody needs the risky option. That is a large part of why businesses come to us.

If your staff are using AI, you need one, and they are, whether or not it has been approved.

It does not need to be long. It needs to say which tools are approved, what must never be pasted into them, and who to ask when someone is not sure. We will help you write one as part of any engagement.

No. What we build takes off your team the work a human should never have been doing, so they get back to the parts that need a person: the clients, the crew, the judgement calls.

Your people stay. The noise goes.

That is the part most AI projects get wrong, so we plan for it rather than hoping.

Three things make the difference. It goes where people already work, so there is no new system to remember. It is introduced to a small group first, so the early problems get found before the whole company sees it. And the instructions live inside the tool itself, so you can ask it what it can do for you. That works better than a training session people forget by Thursday.

We start by asking what they are worried about, before showing anything.

Most of what people know about AI comes from headlines, and most of the fear is reasonable. A first session that opens with questions rather than a demo gets further than one that opens with a screen share. The questions people ask also tell us where the real friction in the business is.

A conversation, and it costs nothing.

After that, an on-site audit. It starts with a survey of your individual staff, because the people doing the work know where it goes wrong. Then we walk your processes ourselves. You get back a ranked list of opportunities, scored on impact against effort, so you can see what is worth doing first and what can wait.

Smaller builds go live in weeks. Bigger systems take a few months.

As a concrete example, a knowledge agent is roughly a couple of weeks to build, plus a week of testing with a small group before it goes wider. Most clients feel it from the first workflow we switch on, in hours handed back or revenue found.

Less than people expect. Three things.

One named contact for day-to-day decisions. An introduction to whoever looks after your IT, so we can arrange access, and we will explain to them what we need and why. And your documents: handbooks, policies, procedures, price lists, whatever your business knows but has never had in one place.

During the build, adding or changing documents is included. There is no charge for feeding the thing.

No, and most businesses we work with are not. Part of our job is getting the foundations right first: communication, process, documentation. Then the AI goes on top.

Otherwise you are just automating chaos.

Then we will tell you, and we have told people before.

Sometimes an audit finds the problem is not a missing AI layer. It is that quoting lives in Word, pricing lives in a spreadsheet, jobs live in Trello, and nothing talks to anything. Automating on top of that just makes the mess faster. In those cases the right first move is a proper system underneath, and we will help you choose and roll one out before we build anything of our own on top.

Usually, and we check before we promise.

Most modern systems, Xero, monday.com, Microsoft 365, Google Workspace, can be connected to. Some New Zealand industry software cannot, or only partly, because it was never built to be connected to. That is a real constraint and it is the first thing we investigate, because it changes what is possible and what it costs.

Where a system genuinely cannot be connected, there are usually workarounds. We will tell you honestly which situation you are in, rather than discovering it halfway through a build.

No. Builds are managed with regular check-ins while in progress, and support continues afterwards: weekdays, with a response within one business day and urgent issues prioritised.

Urgent means something materially affecting your business, not a new idea. New ideas are welcome, they just get scheduled rather than dropped in front of everything else.

Yes, and it is a genuinely good place to start. We come to you, get your team up to speed on what this stuff can actually do, and surface the opportunities hiding in your week.

Workshops are not the end game though. The real value is in what we build from what we find.

Once the engagement has ended and everything is settled, we hand over the automations, workflows and associated accounts, and support a proper handover so nothing stops running.

Because handover is real technical work, a transfer fee applies. It is quoted on the scope of what needs moving, so you will know the figure before you decide. No surprises at the door.

It is covered in the agreement, and it is a fair thing to ask a small company.

If we were no longer able to operate, we would deploy your systems onto your own infrastructure or hand them to a provider of your choosing. The protection is written down rather than promised, which is deliberate. An agreement that only works while everyone is healthy is not much of an agreement.

No. Retainers run on an initial term with 30 days written notice to end, and get reviewed together at the end of each term rather than rolling on by default.