Straight Answers

FAQ

Everything business owners ask before trusting their first agent.

What exactly is an AI agent?

An agent is software that doesn't just answer questions, it takes actions. It reads your inbox, decides what a reply should say, sends it, updates your CRM and books the follow-up. Think of it as a digital team member that works 24/7 and never forgets a step.

Is my business big enough for AI agents?

If you get customer enquiries, send quotes or chase paperwork, you're big enough. Our smallest clients are solo tradies running one reply agent; the largest run coordinated agent teams across sales, service and operations.

Where does my data go? Is it safe?

Your data stays in systems you control wherever possible, and agent providers are configured with strict data-retention and no-training settings. Sensitive actions (payments, contracts, anything customer-facing initially) require human approval. We'll walk you through exactly where every piece of data flows before anything goes live.

How long does a build take?

Most first agents are live within two to four weeks. Reply and qualification agents are at the fast end; voice agents and deep integrations take the longest because of testing with your real call flow.

What tools do agents work with?

Email (Gmail, Outlook), calendars, CRMs (HubSpot, Salesforce, Twenty), job management (Servicem8, simPRO, AroFlo), practice software, Xero, MYOB, website forms, SMS and phone systems. If it has an API, or even just an inbox, we can usually connect it.

What does it cost?

Builds are fixed-price per agent, quoted after the free scoping call, with a monthly plan covering hosting, monitoring and improvements. Most clients find their first agent costs less per week than the admin hours it gives back.

What happens when the agent is unsure?

It stops and asks a human. Confidence thresholds, approval queues and full decision logs are built into every deployment. Agents earn autonomy gradually, they start by drafting, and take over routine actions only once their accuracy is proven on your real workload.

Can agents make mistakes that hurt my customers?

Any system can make mistakes, which is why agents ship with guardrails: approval steps for sensitive actions, tone limits, rate limits and rollback logs. In practice, most errors we catch are the kind your busiest human team member makes on a Friday afternoon, except agents make them at 3am and correct themselves.

We already use [software], why not just use its built-in AI?

Built-in AI features are generic and live inside one tool. Your business runs across five. Custom agents connect the whole picture and are trained on your customers, your pricing and your way of doing things, not a template shared with every other subscriber.

How do we get started?

Book the free scoping call. Thirty minutes, no slides: we map your workflows, find the agent with the fastest payback, and you leave with a plan, whether or not you build with us.