I grew up in Scotland and went to a Harry Potter-like boarding school where I fell in love with mathematics. I was following the path my father imagined would lead me into a desk job, but instead, when I left school, I became an apprentice carpenter.

That decision ultimately took me from working as a journeyman carpenter and high-rise steel worker, where I struggled to find good, consistent work despite having valuable craft skills, to becoming a general contractor, where I experienced the daily frustration of trying to find skilled workers like myself. So same problem, just in reverse.

With 96% of contractors today saying they can’t find the workers they need, the industry has spent decades assuming the problem is a labor shortage, and therefore that the solution is to build machines or train more people.

That assumption is now driving billions of dollars of investment into physical AI and training more skilled workers, including major commitments from Meta, Google, BlackRock, Lowe’s and others.

But my own unlikely and rather winding path made me question the underlying assumption that the primary constraint is a shortage of labor. I had been the skilled worker the industry desperately needed but couldn’t see, and later the employer desperately searching for workers I knew were out there. Perhaps because of my passion for mathematics, I started looking at the numbers and came to a very different realization that only 1 in 20 workers are actually missing from the labor pool.

Construction, it turns out, isn’t fundamentally constrained by a shortage of workers, but by a lack of digital access to the 95% of workers it already has.

That insight became the foundation for Skillit.

From Local Sourcing to National Routing

Today, as a historic infrastructure buildout unfolds, skilled labor has become the biggest constraint on economic growth and is making this access problem much more urgent.

A hyperscale data center in Dallas may suddenly need hundreds of electricians. A semiconductor fab in Phoenix needs another combination of trades. An advanced manufacturing project in Tennessee needs another.

We know from our biggest enterprise customers today that these projects don’t rely on where the workers happen to live but on whether the right skills can reach the right project at the right time.

That’s why construction is beginning to move from locally sourcing labor to nationally routing it, and that requires a very different kind of labor infrastructure.

Workers need to become digitally discoverable through a deep understanding of their skills, licenses, project experience, preferences and availability, not simply a résumé. As a craft worker, I knew my hands were at a premium, so we made onboarding as effortless as possible by empowering workers to simply talk. Skillit’s voice AI then turns natural conversation into rich, structured craft data in real time, capturing nuance a résumé never could.

Employers’ live demand is understood with the same precision. Skillit then matches that demand against proprietary worker intelligence, resolves the constraints between workers and work, and intelligently routes people to the opportunities where they are most valuable.

We built Skillit around that idea.

Making Connections – Changing Lives

More than 230,000 craft workers have now signed up, and over half of the network is concentrated in five of the trades most critical to major construction projects: electricians, carpenters, heavy equipment operators, concrete workers and general laborers. 

The results are also beginning to show what happens when access improves. Skillit is now used by 13 of America’s top 25 data center contractors, including all five of the top five, and our connection rate between employers and workers is already seven times higher than Indeed’s. 

But if I’m honest, the metric I care about most is what happens to individual workers.

One welder in Dallas, Eric, came to Skillit after bouncing between short-term jobs looking for something more stable. We routed him to multiple employers. He chose Mortenson because of its safety record, benefits, nearby work and workforce development program. Today he earns 25% more, has greater job security and was able to put a down payment on a home for the first time.

Eric wasn’t missing from America’s labor force. He was there all along—it's just that the  infrastructure connecting his skills to the right opportunity was missing.

As America races to build the physical infrastructure behind AI, energy and advanced manufacturing, training more people will absolutely matter, as will physical AI. But both take years to make an impact at the scale we need. And so, while we spend billions training the next generation of workers and building robots, we will be wise to get much better at finding, understanding and coordinating the millions of skilled workers who already exist.

To learn more visit skillit.com.