Civic software we build and run

Four public-interest tools we designed, built and still operate ourselves. No client commissioned them, none of them sells anything, and every claim they make traces to an official record.

The portfolio

Why a consultancy builds these

These four are not client work and not marketing. They exist for three reasons, and it is worth being plain about all three.

They are the hardest engineering we get to do. A client project has a scope and a sign-off. A public tool that states the law has neither: it has to be right when nobody is watching, stay right as the underlying records change, and say so plainly when the records cannot answer. That constraint produced the claims register in FreeCamps, the provenance grading in Public Trails, the neutrality lint in how2vote and the suppression handling in Surgery Waits - techniques our engineers now bring to commercial work. There is no faster way to develop a team than giving them a problem where "probably fine" is not an acceptable answer.

They are useful to people who cannot commission software. Someone comparing surgery waiting times across a border, or working out where they can lawfully sleep tonight, is not a market. Each of these sites is free, carries no advertising, no lead capture and no link to anything we sell, and each publishes the licence of every dataset it uses. That is a deliberate limit: the moment one of them becomes a funnel it stops being trustworthy, and its usefulness is the whole asset.

They are how we practise the parts of delivery that are hardest to demonstrate. Compliance, provenance and public accountability are easy to claim in a capability statement and hard to show. Here they are inspectable. how2vote carries an electoral authorisation on every screen, a published control register and two named accountable signatories; FreeCamps holds 86 legal claims to a source, an exact quote and a re-verification deadline, enforced by the build; Public Trails refuses to publish a route it cannot evidence. A prospective client can read the code.

Who runs them

Cameron Young, our CEO, brings a legal background to work that turns statutory material into software - which is why these products treat a legal claim as something with a source, a quote and an expiry date rather than a sentence in a template. Tara Buckley, our COO, brings experience working with government and not-for-profit organisations, which shapes what these tools refuse to do: no personal data collected where it is not needed, no crowd-sourced claim dressed as an official one, and an obligation named on the page rather than buried in a policy.

Both are named accountable signatories on how2vote's compliance register, and its Privacy Act determination records a two-person review rather than a single sign-off. Naming the people who are answerable is part of the design.

The common method

All four are built the same way, and it is the same way we build for clients: official sources only, every claim traced and dated, unknown rendered as unknown, and the guarantee enforced by a gate in continuous integration rather than by anyone remembering. Each publishes the licence of every dataset it uses, and how2vote is open source under AGPL-3.0 with the rest to follow. If that is the standard your project needs, see how we approach custom software development and platform engineering.

The four, and how each one was built

Questions people ask about these projects

Who funds these, and what do they cost to use?
National Digital funds them and they are free to use. There is no advertising, no subscription, no account and no lead capture on any of the four, and none of them links to anything we sell. That is a constraint we hold deliberately: a public tool that quietly funnels its readers somewhere commercial cannot make the neutrality and provenance claims these tools make, and those claims are the entire value.
Why spend engineering time on work nobody is paying for?
Because it is the most demanding work we do and the most useful thing we can show. A public tool that states the law has to be correct unattended, stay correct as its sources change, and decline to answer when the records cannot support one. Building to that standard produced techniques - claim registers, provenance grading, gates that fail a build on an unsourced statement - that our engineers now apply to client platforms.
Can I see the code and reuse the data?
The data licences, yes, on every site. The source, so far, for how2vote: it is AGPL-3.0 with public architecture decision records and a governance document, and the others are intended to follow. The government records underneath carry their own licences from their own publishers and are never sub-licensed by us - where a publisher has set no licence, the data is withheld rather than republished.
Who is accountable for what these tools say?
National Digital, named on every page as the publisher, with two named accountable signatories on the compliance register behind how2vote. Every legal claim carries its source, the exact quote and the date it was retrieved, so a reader can check the claim rather than trust the publisher. Where a tool cannot support an answer it says unknown, and unknown is never rendered as permission.
Can you build something like this for a government or not-for-profit body?
Yes, and the four sites are the working demonstration rather than a capability statement. The same method applies: official sources only, provenance on every claim, refusal where the records cannot support an answer, and the guarantee enforced by a gate in continuous integration rather than by review discipline. Get in touch and we will tell you what we would do first.

Building something the public has to be able to trust?

Tell us what you're working on. We'll come back within one business day with what we'd do first and what it would take.

or enquiries@nationaldigital.com.au1300 800 855