Construction has long been one of the most data-rich yet data-poor industries going around. A single commercial project can generate tens of thousands of drawings, models, clash reports, requests for information and site photographs, and yet the people making decisions on the ground routinely struggle to find the one number or the one revision that actually matters. That gap between how much information a build produces and how little of it is usable in the moment is exactly the space that construction technology firm Revizto is now trying to close with artificial intelligence.
The company has announced what it describes as an open AI layer for construction data, a system intended to sit across the coordination and building information modelling (BIM) work that project teams already do inside its platform. According to SecurityBrief Australia, the emphasis is on the word open: rather than locking intelligence inside one closed product, the layer is pitched as something that can work across the mix of tools and data sources that a typical project already juggles.
Why an open layer, and why now
To understand the significance, it helps to know what Revizto does. The platform is used by architects, engineers and contractors to bring their separate 3D models together into a single coordinated environment, to spot clashes between, say, a ventilation duct and a structural beam before either is built, and to track the issues that flow from that coordination. It sits in the category the industry calls common data environments, the shared digital space where a project’s information is supposed to live.
The problem is that even inside a well-run common data environment, the intelligence is largely manual. A coordinator still has to open the model, run the clash detection, read through the results and decide what is urgent. An open AI layer aims to change the nature of that interaction. Instead of hunting through a model tree, a project engineer could in principle ask a plain-language question about outstanding issues, overdue items or where a particular clash keeps recurring, and get an answer drawn from the underlying data.
The decision to make the layer open rather than proprietary is the more interesting strategic call. Construction runs on a patchwork of software, and no single vendor owns the whole workflow. A design consultancy might model in one package, the builder might estimate in another, and the client might demand deliverables in a neutral format such as IFC, the open standard for exchanging building models. An AI layer that only understood one vendor’s data would be of limited use on a real site. By positioning the layer as open, Revizto is signalling that it wants to be the intelligence that reads across those sources, not just its own.
Two ways to read the move
The optimistic view is that this is precisely the kind of applied AI the built environment needs. Construction productivity has been notoriously flat for decades, and rework driven by coordination errors is one of the biggest hidden costs on any large job. If an AI layer can surface the clash that would have become a costly variation, or flag the RFI that has been sitting unanswered for three weeks, the value is concrete and measurable rather than speculative. This is not AI writing marketing copy; it is AI pointed at a problem that already has a dollar figure attached to it.
The more cautious reading is that “open AI layer” is a phrase doing a lot of work, and the detail matters enormously. Openness can mean many things, from genuine interoperability with third-party tools through to a lighter arrangement that still keeps the valuable data inside one ecosystem. There is also the perennial question of trust. Construction teams are accountable for buildings that people live and work in, and an AI that confidently summarises the wrong revision or misreads a model is not a harmless error. Any tool of this kind will need to show its working, point back to the source drawing or issue, and let a human verify before anyone acts on it. The industry’s appetite for a black box that simply asserts answers is, rightly, close to zero.
The Australian stakes
For Australia, this is not an abstract international product story. The country is in the middle of an infrastructure and construction pipeline worth hundreds of billions of dollars, spanning transport megaprojects, hospitals, social housing and the data centres now being built to power the AI boom itself. Governments here have been steadily pushing BIM and digital engineering requirements into major project procurement, particularly in New South Wales and Victoria, which means the digital coordination work that Revizto’s layer targets is already mandated on a growing share of public jobs.
At the same time, the local sector is under acute pressure. Insolvencies among Australian builders have run at painful levels, margins are thin, and skilled coordination staff are scarce and expensive. Tools that let a smaller team manage the same volume of design information, or that catch expensive errors earlier, land in a market that is genuinely motivated to adopt them. That combination of digital mandates and commercial stress makes Australia a more receptive testing ground for construction AI than many markets overseas.
There is a data-sovereignty dimension too. Project information on a defence facility, a hospital or critical infrastructure is sensitive, and Australian clients are increasingly particular about where that data is processed and stored. The fact that this news reached local readers through a security-focused outlet is a reminder that, for many buyers here, the first questions about any AI layer will be about governance and residency, not just features. Any vendor courting the Australian market will need clear answers on where the models run and how client data is handled.
What is next
The real test will be adoption on live projects rather than the announcement itself. Watch for whether major Australian contractors and engineering consultancies begin naming this kind of AI capability in their digital engineering approaches, and whether the promised openness translates into genuine connections with the other tools on a project. Equally telling will be how the layer handles accountability: whether it consistently cites its sources and defers to human judgement on anything that carries safety or contractual weight. If it clears those bars, construction may quietly become one of the more convincing places that enterprise AI earns its keep in Australia.
Sources: SecurityBrief Australia.
