Revizto opens model data to AI workflows

Revizto has expanded access to building model and project data, allowing construction companies to connect live information with external analytics, enterprise systems, and artificial intelligence workflows.

New model and object property endpoints provide read only access to detailed metadata without requiring users to open the main application. External systems can retrieve object type, material, system, quantity, author, timestamps, and custom properties.

Business intelligence dashboards, cost estimation, procurement, facilities management, carbon analysis, progress monitoring, and structured handover can all draw on the information. Automated pipelines should reduce reliance on repeated manual model exports.

Revizto’s connected project environment also includes interfaces for issues, clashes, two dimensional information, users, and project administration. Its Model Context Protocol capability allows approved AI tools to query project information while retaining permissions applied within the source platform.

Potential integrations include enterprise conversational AI, internally hosted models, digital assistants, reporting tools, and specialist construction applications. Current project information can therefore be made available to authorised workflows without creating a separate uncontrolled copy for every use.

Early users have applied the model data interface to installation progress dashboards, material quantities, carbon calculations, procurement processes, model auditing, and facilities management handover.

Mo Shana’a, chief executive and cofounder of Morta, said: “For too long, model data has been out of reach for the people controlling delivery — from progress monitoring through cost control to handover.”

He said connecting Revizto data to the processes already used by project teams could turn the model from a static deliverable into an everyday operational resource.

Omar Magdi Eid Awad, digital construction coordinator at ALEC, said the interface had enabled an automated pipeline connecting Revizto with other project platforms. Direct access to model and object properties, he added, completed an important part of the wider API.

The expansion follows growing concern about data ownership and control within construction technology. Revizto’s industry research found that most construction chief information officers were concerned about ownership across increasingly complex software environments.

AI depends on governed project information

Construction companies are testing AI for document search, reporting, risk identification, design review, cost analysis, and programme support. The reliability of those outputs depends on whether the underlying project information is current, structured, and accessible.

A language model cannot reliably establish which drawing is approved when file naming is inconsistent, superseded revisions remain active, and issue histories are incomplete. Automation can process poor information faster without resolving the ambiguity behind it.

Direct access to live platform data reduces part of that risk because an external workflow can query the current system instead of relying on a spreadsheet or model exported several days earlier. The result still depends on the quality of the source record.

Permissions are particularly important on construction projects because multiple organisations create information under different contractual responsibilities. A subcontractor may need access to selected coordination issues but not commercial records, while a facilities team may require asset data without visibility of unresolved tender information.

Any AI connection must respect those boundaries. Broad access granted for convenience can expose commercially sensitive, personal, security related, or legally privileged material beyond authorised users.

Audit trails are equally necessary because companies need to know which system accessed data, what information was returned, how it was used, and whether a human reviewed the resulting decision or document.

Reading model quantities and properties can support procurement and carbon reporting, although those figures require context. A model may contain duplicate objects, design allowances, temporary components, or elements that have not reached a suitable level of definition for commercial use.

Automated extraction must therefore remain connected to status, revision, classification, and responsibility. Without those controls, a live interface can give false precision to information that remains provisional.

Handover presents another substantial opportunity because asset data can move from design and construction systems into facilities management processes without repeated manual entry. That continuity can reduce the loss of information that often occurs near completion.

Asset information requirements must still be agreed early. Opening access at the end of a project will not repair missing serial numbers, inconsistent classifications, or incomplete commissioning records accumulated during delivery.

APIs also allow companies to avoid placing every function inside one platform. Specialist applications can remain in use while common data feeds into reporting, commercial, operational, and decision making processes.

Each connection nevertheless requires maintenance as software changes, credentials expire, project teams move, and data structures evolve. Poorly governed integrations can become another source of hidden operational risk.

Revizto’s expanded interfaces provide the technical route for broader project data use. Companies will gain most where the connection is supported by disciplined information management, controlled permissions, and clear human responsibility for decisions produced through automated workflows.