Four ways we build software that fits how your business actually runs.
Some engagements start with a website that undersells the business behind it. Others start with a spreadsheet that has quietly become critical infrastructure, a repetitive task nobody has had time to automate, or a product idea that has outgrown a prototype. These four capabilities cover that whole range — and they're rarely used in isolation. A website often surfaces the business system behind it; a business system often reveals where automation would help.
Websites
Business websites that load fast, communicate clearly, and are straightforward to maintain after launch.
Business systems
Custom CRMs, quoting tools, portals and scheduling systems shaped around how your team actually works.
Automation & AI
Controlled automation and AI-assisted workflows that remove manual admin — with a person approving anything that matters.
Digital products
SaaS platforms and vertical software, taken from a proof of concept through to production and ongoing iteration.
A website is often the only proof a prospect has that the business is legitimate
A meaningful share of small and mid-sized businesses are still represented by a slow template site, a page nobody has updated in years, or a builder platform that fights against basic content changes. For a considered purchase, the website is frequently the entire interaction a prospective customer has with the business before deciding whether to trust it.
All of State Zero's delivered client work to date sits in this category — sites built for a Brisbane gaming PC retailer, a trauma-informed mental health practice, and an independent Thermomix consultant. See the work .
The parts of the business that don't show up on the website
Quoting, job tracking, approvals and reporting often live across a spreadsheet, an inbox and whatever system was bought under pressure. Nobody has the full picture, and someone is holding it together by hand.
AI is a capability we apply carefully — not a pitch
Most of what businesses call "automation" doesn't need AI at all — it needs a reliable rule. Where AI does earn its place, we're precise about which of these three tiers a problem actually needs.
Rules-based automation
Deterministic steps that fire the same way every time — no model involved, and nothing to second-guess.
A grounded knowledge assistant
Answers drawn only from your own documents and procedures, not the open internet, so it says "I don’t know" rather than guess.
An agentic workflow
An "agent" is software that can take a multi-step action on your behalf — drafting a reply, updating a record. It prepares the action; a person still approves it.
Some ideas need software of their own
Sometimes the right answer isn't a website or an internal tool — it's a product in its own right: something customers or members log into, pay for, or depend on. We take these from a proof of concept through to something that can run in production, and keep shaping it once real usage starts.
How an engagement begins
Most engagements start with a conversation, not a scope document. We talk through what isn't working, then work out together where the highest-value starting point is.
A conversation
You tell us what is not working. We ask questions until the problem is specific.
Discovery, if needed
For larger or less-defined problems, a short paid phase shapes scope before anything is built.
Build in increments
We ship working pieces, put them in front of real use, and refine from there.
Not sure which of these fits?
That's a normal starting point. Tell us what the business is dealing with and we'll help work out which of these — or which combination — actually applies.