I keep coming back to one line from a McKinsey piece on agentic AI in construction: the competitive moat in this industry's AI future belongs to the firms that control proprietary project data, the workflows where decisions get made, and the ability to charge for outcomes rather than hours.
I shared that line on LinkedIn a few weeks ago, and it's still one of the most-engaged posts I've written this year. But the honest reaction I had before I ever hit “post” was simpler: we've been building that moat for years. We just never called it that.
It's easy to think of AI as something happening on a screen. But it's showing up in steel and concrete now, too — new power partnerships, new data centers, states debating who pays for the infrastructure behind all of it. AI is moving from experimentation into real capital investment. Which means the same discipline it takes to plan, fund and manage a physical asset is exactly what owners need to be applying to something less visible: their own project data.
Here's where it gets concrete for us. Customers are starting to come to VPO and ask us to do something we didn't originally build the product to do: take their historical project data — RFIs, change orders, meeting minutes, daily reports — and use AI to help them find their own lessons learned.
That request tells you everything. Most of that data gets generated, filed away, and never looked at again — until something goes wrong and it ends up in litigation. But it's full of insight. It's the record of what worked, what didn't, and what quietly cost somebody money the last time around. Every project should teach the next one. Most owners just don't have an easy way to let it.
This week, we kicked off a partnership with Carnegie Mellon's Master of Software Engineering program — consistently ranked among the best software engineering graduate programs in the country, and it's right here in our backyard in Pittsburgh. VPO was selected as one of this fall's capstone projects. A dedicated team of CMU students will spend the semester digging into project data to identify exactly which insights can be automated: what patterns are worth surfacing, and what “lessons learned” could look like when a computer helps you find them instead of a person combing through years of files.
We don't have all the answers yet. That's the point of a capstone. But it's a real, active step toward turning almost two decades of owner-side project history into something an owner can actually use — not just store.
None of this works if the underlying data isn't yours to begin with. As a colleague with 50-plus years in construction put it to me recently: AI doesn't replace judgment. It replaces the three days it used to take to build a spreadsheet. That's a real gain — but only if what's feeding the AI is data you control, built from your own projects, not someone else's.
That's exactly what we're digging into at our webinar on September 30th — Data Ownership in the Age of AI: What Every Construction Owner Needs to Know. Who controls the information your projects generate. Whether you'll still be able to access and understand it years from now. And what it takes to make sure your project history is working for you, not just sitting in a folder somewhere.
I hope you'll join us.
And here's to Building Better, Together.
Laura Nee
CEO & Founder, VPO
Webinar — Data Ownership in the Age of AI: What Every Construction Owner Needs to Know
And as AI becomes more deeply embedded in capital project management, the same principle will matter even more. The usefulness of AI will depend not just on how much information an organization has, but on whether that information captures developing conditions, decisions and outcomes in a way that remains understandable over time.
On September 30, VPO will continue the conversation by looking at how the information owners create and control today will shape what they can do with AI tomorrow.
Want to see how VPO gives construction owners real-time visibility into project health — across every issue, every risk, and every project in your portfolio? Let's show you.