17 March 2026
MCP and the connected construction office
Every so often a piece of technical plumbing changes what is practically possible without ever becoming famous. In the AI world, the Model Context Protocol, MCP, is that piece of plumbing, and it matters more to a construction business than most of what makes the headlines.
You do not need to understand the engineering. You need to understand what it changes.
The problem it solves
Until recently, an AI assistant was a very well-read colleague locked in a room with no phone. It could reason about anything you pasted in, but it could not look anything up in your systems or do anything on its own. Every integration between an AI and a tool, SharePoint, Outlook, a cost system, a CDE, had to be custom-built, one connection at a time. Expensive, fragile, slow.
MCP is a standard way for AI models to connect to tools and data sources. Think of it like the USB standard: any tool that speaks it can plug into any AI that speaks it. Instead of custom wiring for every pairing, there is one socket shape. The assistant can now, with permission, read the folder, search the inbox, query the tracker, and write the update back.
That is the shift: from an AI that answers questions about what you paste in, to an AI that can operate across the systems your business already runs on.
What it looks like on the ground
Take the workflows we write about constantly. A drawing arrives by email. A connected assistant reads the transmittal, files the drawing to the right SharePoint location with the right name, updates the drawing register, checks whether the revision supersedes anything in the current package, and drafts the notification to the affected trades. Six systems touched. No human clicking between them.
Or the Monday report. The assistant pulls progress from the site record, costs from the tracker, programme status from the planner's export, opens last week's report as the template, and produces this week's draft with the changes highlighted. Your PM edits judgement, not formatting.
None of this was impossible before. It was just bespoke every time. A standard socket collapses the cost of connecting things, which means the automation that used to make sense only for a Tier 1 budget now makes sense for a regional contractor.
The caution that comes with it
An AI that can reach into your systems is powerful in both directions, which is why the guardrails matter more, not less. Access should be scoped to exactly what each workflow needs. Every action should be logged and auditable. Anything that leaves the business or carries contractual weight goes through a person. And the whole arrangement should live in your environment, under your identity controls, not on someone's personal account.
The standard is still maturing, and it is not the only way to connect systems; conventional APIs and automation platforms still carry plenty of the load. The direction, though, is set: the walls between your AI tools and your business systems are coming down.
Why this favours the prepared
Connected AI amplifies whatever it is connected to. A clean SharePoint, consistent naming, a structured project record: the assistant flies. A decade of chaos: the assistant faithfully retrieves chaos, faster than ever.
Which is why our advice has not changed since before any of this existed. Sort the data foundations, capture the records, structure the information. Every improvement you make there now pays twice: once immediately, and again for every connected workflow you switch on later. The plumbing has arrived. The businesses whose information is ready for it will feel like they got a head start, because they did.