/ Why now

Why now

The largest build-out in American history needs people first.

Between 2025 and 2032, U.S. companies are projected to put $10.3 trillion into AI infrastructure. Data centers. Power. Networks. Chips.

That is 3.6% of GDP, every year, for eight years. More than the railroads. More than the highways. More than the internet. And not one dollar of it builds anything until somebody hires the people who do the building.

$10.3T

projected U.S. AI infrastructure investment, 2025 to 2032. Data centers, power, networks, chips

3.6%

of U.S. GDP, every year. Larger than any infrastructure boom in the country's history

1.6×

the railroad boom's share of GDP. The previous record, set in the 1870s and 1880s

The scale of it

Canals. Railroads. Highways. Fiber. Then this.

Every generation gets one or two build-outs that reshape the economy. Measured against the size of the economy at the time, nothing in American history comes close to what's being built right now.

Every great American build-out, side by side

Average annual investment, % of GDP

Canals1836 to 1841

0.66%

Railroads1870 to 1890

2.24%

Electrification1905 to 1925

0.50%

Highways1956 to 1973

1.13%

Telecom and fiber1996 to 2003

1.10%

AI build-out2025 to 2032

3.63%

Source: Stijn Van Nieuwerburgh, Columbia Business School, "Financing the AI Buildout," Brookings Papers on Economic Activity, Fall 2026. A projection, not a guarantee.

Where the money goes

Four places the money lands. Every one of them is a hiring problem.

The $10.3 trillion isn't one thing. It breaks into four kinds of infrastructure, and each one needs a different kind of person that most companies can't find on their own. Those people are who you'd be finding.

Data Center & Critical Infrastructure

The buildings.

Data centers the size of small towns, going up faster than anyone has built them before. Someone has to choose the site, design it, build it and keep it running.

Who gets hiredMechanical and electrical engineers, construction and program leaders, critical facilities operators

Energy Transition & Industrial

The power.

AI runs on electricity, and the grid wasn't built for this much of it. New generation, new transmission, and new ways to keep it all stable.

Who gets hiredPower systems engineers, energy project developers, grid and utility-scale leaders

Enterprise Software & Cloud

The networks.

Thousands of chips have to talk to each other like one machine. The networking and infrastructure software underneath is its own discipline.

Who gets hiredNetwork architects, distributed systems engineers, infrastructure and platform leaders

Semiconductor & Microelectronics

The chips.

The most sought-after silicon ever made, and the equipment that makes it. Some of the rarest engineering talent anywhere.

Who gets hiredChip design and verification engineers, packaging and process specialists, hardware leaders

And on top of all of it, the AI and machine learning teams that turn the infrastructure into products. That's where we started, and it's still home.

Picks and shovels

Every gold rush has two kinds of people. We chose the side that always gets paid.

In 1849, most people who went west to dig never struck it rich. Many of the ones who did well were selling what every miner needed: picks, shovels, denim, supplies. They didn't have to guess which claim would hit. They sold to all of them.

This rush is no different. Companies are betting billions on which models, which chips and which data centers win. We don't have to bet. Every one of them needs people, and people are the scarcest thing in the whole build-out.

The miners

Bet everything on one claim
Win big or go home broke
Need the rush to break their way
Compete with every other miner

The picks and shovels

Sell to every claim at once
Get paid whether or not one strikes
Benefit from every new company that shows up
Build relationships with all of them

The scarcest shovel in this gold rush is a person. That's what we sell.

The engineer who designs the chip. The one who powers the data center. The one who trains the model. You'd be the one who finds them, convinces them, and closes the deal.

The obvious question

"Isn't this a bubble?" Maybe parts of it.

We'd rather say it before you Google it. The economist behind the $10.3 trillion number warns the build-out could overshoot. Railroads and telecom both had busts. His paper estimates the industry will need about $3.7 trillion a year in revenue by 2032 to pay for all of it. Nobody knows if it gets there.

Here's how we think about it.

We don't own the bet

Picks and shovels is the hedged position. We get paid when companies hire, not when one of them wins. We don't need to pick the winner.

We're not one market

We recruit across 10+ verticals, from venture-backed startups to the enterprise. AI infrastructure is the biggest story right now. It isn't the only thing we do.

The track still gets laid

The railroad boom had busts. America still ended up with the railroads. Much of the fiber laid before the telecom bust went on to carry the internet's growth. Build-outs leave things behind, and the people who built them.

Watch

The biggest bet in U.S. history, in one segment.

$10.3TThe AI build-out, explained

Fox Business coverage of the Brookings projection.Click to play

What this means for you

Books get built during build-outs.

The whole model on our sales career page rests on one idea: the relationships you open early keep paying for years. The best time to open them is when an industry is forming, not after it's settled.

That's right now. The engineering leader running a team of twelve at a data center startup this year could be running a few hundred people in five years. If you're the one who helped build that team, you're the first call for the next one.

You won't get another decade like this one. The question is whether you spend it selling software or building the companies.

See the open role → How the sales career works

One conversation. Then decide what you want to build.

Leave your details. Someone who has run these desks will reach out.