Backlogs now exceed eight months across construction, and that timeline is even longer for data centers at more than 11 months. Even with 26,000 jobs added in March, hiring alone isn’t enough to keep up.
The industry doesn’t only need more workers—it needs a new approach to delivering projects.
Rethinking How the Industry Views AI and Software
If a company spends $1 million on a piece of machinery, say a large bulldozer or a heavy-duty mining truck, no one bats an eyelash. It’s an essential component for the business. With good care, that equipment will last for years and help the business earn back its initial spend.
Meanwhile, if someone spends $1 million on software, there will be some raised eyebrows. Yet are those scenarios really that different?
In both cases, the investment is in something that’s helping people get things done more effectively. There needs to be a fundamental mindset shift that AI, especially agentic AI, is not actually a technology cost. Rather, it’s a labor cost.
Companies need to treat these agents like teammates. Giving superintendents and senior project managers their budgets to perform work and then letting them decide how to spend it to get the job done most effectively.
NVIDIA CEO Jensen Huang proposed a similar plan at the company’s annual GPU Technology Conference. He said he plans to pay his engineers with AI tokens, on top of their salary, to enhance their efficiency and output. Huang estimated that, in addition to his 42,000 “biological” employees, he’s going to have “hundreds of thousands of digital employees.”
“Every engineer who has access to tokens will be more productive,” he said.
Applying This Mindset to Resource Management
Construction leaders wouldn’t send a crew to a jobsite without the right tools and equipment—that’s simply not how the work gets done. The same thinking needs to apply to information. Labor, equipment and materials have always been the core resources in construction, but today, data is just as critical. When that data is fragmented across systems and teams, field leaders are effectively being asked to build with one hand tied behind their backs.
Giving field teams what they need to perform means fully integrating data across labor, equipment, materials, schedules, budgets and project documentation. Not as a technology initiative—but as a commitment to helping people do their best work. AI agents are a key part of that. Think of them the way any other resource in the field is approached: they have a job to do, and the return on that investment shows up in productivity, schedule performance and margin.
Imagine a superintendent who can look across projects simultaneously—drawing from crew assignments, equipment availability and productivity data—to put the right resources in the right place, not just to meet the schedule, but to beat it. In an industry facing real labor constraints, how the industry deploys the workforce matters just as much as how many people it has. Agents help maximize the return on every worker in the field.
And the case for this isn’t complicated. Ask anyone in construction what holds them back, and they’ll say it’s visibility—not knowing what they need to know, when they need to know it. The volume of project data today has outpaced what any team can manually manage. AI agents tackle that volume, surface what matters and help teams think through scenarios before problems become crises. That’s not a luxury. That’s the cost of building well.
Building a More Efficient Foundation
Procore was recently in communication with an executive at a very large construction company, who’s been in this industry for 40 years and has seen nearly everything there is to see. The Procore team asked him, “Why do we bill on a monthly basis?”
He didn’t have an answer. He simply said that he’d never thought about that before.
And, for most, the reason is because that is the way it has always been done. Perhaps the reason it has always been done that way is that the process is such a burden.
But what if those processes weren’t so burdensome? What if, instead, the industry could automatically validate progress? Eliminate the need for lien waivers? Digitally sign off on these things?
That’s just one example of why “it’s how it’s always been done” doesn’t fly anymore. The industry is moving too fast to maintain status quo, and the innovative companies will be the ones rising to the top.
Consider data center owners: Money is generally not an issue for them. They’re willing to fund large-scale projects, but they want the job done fast, often faster than what’s realistic for humans alone. If that team of humans has an assist from an AI agent crew, the path to success is easy to visualize.
An AI agent can help predict issues or mistakes before they arise, during planning phases, at jobsites and with owned and rented equipment. Their complementary work saves money in preventing errors and shaves days, weeks, or even months off the project timeline.
To be clear, this concept is not about replacing people. It’s about reducing the cost of human labor and shifting some of those labor costs to AI.
If an AI agent can halve the time it takes to complete a task, freeing up human construction workers to contribute more high-value results to the business, wouldn’t a company want to grab hold of those capabilities?
Construction software is designed to support, not overtake. As a result, it increases the revenue per employee and drives impact across the company.
The industry has to break some paradigms to keep up with demand—in housing, data centers, office buildings and all kinds of other infrastructure. Shifting this mindset is an excellent foundation to take on more volume, do more work and do it faster.
SEE ALSO: DATA CENTER DEBATE: THE EFFECT OF THE RATEPAYER PROTECTION PLEDGE
