Revizto launches open AI layer for construction data

Revizto has launched an Open Data & AI Layer for construction organisations. The offering is aimed at AECO users that want to connect artificial intelligence tools to project data under their own controls.

The launch adds two technical elements to the platform: a read-only Model and Object Properties API, and an MCP Server based on the Model Context Protocol. Together, they are intended to give approved software and AI assistants direct access to project information without moving data outside a customer's environment.

Revizto is positioning the move around a persistent concern in the construction sector: data ownership and the use of AI in regulated or commercially sensitive projects. According to the company's 2026 Bridging the Gap Digital Design and Construction Report, 98 per cent of Australian chief information officers said they were concerned about data ownership and control when choosing technology vendors, while 32 per cent cited regulatory uncertainty as the biggest barrier to extracting value from AI.

The changes come as contractors, designers and owners look for ways to use generative AI and automated reporting tools on live project data without handing information to a single software provider. Revizto says its approach allows customers to choose their own AI systems rather than rely on a built-in proprietary assistant.

The architecture is designed so AI tools connect to where project data already sits. That is intended to help customers keep project intellectual property within their own systems while still using external or internally developed AI applications.

At the centre of the launch is the Model and Object Properties API, which provides programmatic access to model metadata and object-level details, including fire ratings, material specifications and quantities.

The second element is the MCP Server, described as a no-code connector for approved AI assistants. It can be used with tools including Claude, Microsoft Copilot and ChatGPT to query Revizto project data securely.

These additions are supported by a new Developer Portal with application registration and OAuth 2.0 authentication. This gives developers and customers a standard way to build applications and AI-driven workflows on top of Revizto data while retaining governance over access.

Practical use

The new layer can be used to ask natural-language questions about project status, compliance and delivery. Examples include identifying critical fire doors that lack certification, unresolved clashes on a specific level, or assets due for handover that are missing operations and maintenance documentation.

It can also generate project reports and compliance documents from live data, and identify quality issues or incomplete model information across large groups of assets. That reflects a wider shift in digital construction tools from visual coordination towards structured querying and automated checking of project information.

Jason Howden, Chief Innovation Officer at Revizto, said the release is intended to address a practical bottleneck for construction teams dealing with large volumes of information.

"Construction teams already have enormous amounts of valuable project information. The challenge isn't collecting more data; it's turning that information into answers and actions quickly without compromising security. These new capabilities allow customers to ask questions of their project data, automate reporting and connect AI assistants to live project information while keeping complete control over where their data lives and how it's used," said Howden.

Customer response

Early customer comments focused less on AI itself and more on access to structured data. For many large contractors and project teams, a recurring problem in common data environments is the difficulty of extracting object-level information in a consistent format for assurance, reporting and quality review.

Nathan Beplate, Senior Digital Engineering Manager at John Holland Group, said: "Revizto continues to deliver high-quality, well-documented API access, and the introduction of model and object-level endpoints represents a major milestone. It now enables us to perform the full suite of high-value operations we need in a consistent, automated way.

"Having access to these endpoints has been the missing piece for us. It enables true end-to-end data compliance and quality auditing as an automated, scalable process. This is something that was previously very difficult to achieve.

"Direct, structured access to our own data removes a large portion of time-intensive manual interactions - effort we can no longer afford given the scale and pace of delivery," said Beplate.

That emphasis on compliance and audit reflects growing pressure on major projects to demonstrate that digital records, model data and handover information are complete and traceable. Rather than using AI simply as a chatbot, the appeal for contractors may lie in combining structured model data with automated checks and reporting.

Ben Weblin, Solutions Architect at Naylor Love, also highlighted the structure of the data made available through the new interfaces.

"I have no doubt that this data will power great reporting products for us. I have been in the analytics and data space for quite some time and have ingested data from many endpoints across a number of sectors, so I mean it when I say that these APIs are a breath of fresh air. Perfectly structured, logically segmented, properly keyed, and a data dictionary to match," said Weblin.

Revizto, which works with architects, engineers, contractors and owners, said the launch extends its role as an open ecosystem rather than a closed software environment. The move also underlines a broader trend across construction technology, where software suppliers are under pressure to show that customers can retain control of project data while integrating third-party analytics and AI tools.

For project teams, the immediate test is likely to be whether the new access layer reduces manual reporting work and makes it easier to interrogate live coordination data at scale. Early users placed the strongest value on direct, structured access to project information for compliance and reporting.