Selling AI software turned out to be the easy part. Getting it to actually work inside a company is where the money and the headaches live, and Microsoft just made that its next big bet.
On Friday, Microsoft announced a new subsidiary called Frontier Company, backed by $2.5 billion and 6,000 engineers whose entire job is to sit inside customer companies and make AI deployments actually pay off. Judson Althoff, Microsoft's Chief Commercial Officer, called it "the largest, most capable, outcome-driven engineering organization in the industry."
The category has a name: forward deployed engineers, or FDEs. Palantir built its business on them by sending elite engineers to embed with a customer, learn their systems, and wire the software into whatever weird workflow the customer actually runs. Now every major cloud company is copying that playbook because a Copilot license, on its own, isn't landing.
The scale of Microsoft's bet becomes clearer next to what its rivals are doing:
- Microsoft: $2.5 billion, 6,000 embedded engineers, launched as a full subsidiary.
- AWS: $1 billion for its own FDE org focused on agentic AI, announced earlier this year.
- Palantir: Reported Q1 2026 U.S. revenue growth of 104% year over year, with CEO Alex Karp bragging that the company's Rule of 40 score, a common software growth-and-profit yardstick, hit 145%.
Microsoft is outspending AWS by 2.5x on the same problem Palantir has quietly been monetizing for a decade.
The reason the market suddenly cares is that AI pilots keep dying in production. In a survey from workforce analytics firm Orgvue, 92% of organizations said they've invested in AI, but 78% said those projects have either failed outright or gotten stuck in pilot. That's the third year running the number has looked like this. And 83% of the same companies plan to spend more this year anyway.
Oliver Shaw, Orgvue's CEO, put it plainly: "Failed AI deployments are not a technology problem; they're a workforce design problem." Which is another way of saying the model is usually fine. The hard part is everything around it: the data, the process, and the people who need to trust it enough to change how they work.
Even Mark Zuckerberg has been telling Meta employees that AI agents haven't progressed as fast as he'd hoped, which is a striking thing to hear from a company spending tens of billions to build them.
For customers, having engineers from Microsoft, AWS, or Palantir camped inside your building is a mixed blessing. Gary Brantley, the CIO of the NFL, said that AWS's FDEs helped the league launch products like NFL Fantasy AI in weeks instead of months. That's the sales pitch. The quieter reality is that every hour those engineers spend in your stack is an hour spent wiring your future to their models and their infrastructure. Palantir customers already know this. Microsoft's are about to.
Satya Nadella, framing the whole thing internally, said there is "no societal permission for an AI future that eats the intelligence of the companies it's deployed inside." That's a polite way of admitting customers have started asking what exactly they're paying for when the pilot never turns into anything.
The industry spent two years arguing about which model was best, and the answer turns out to matter less than which vendor will send a team of humans to your office to make the thing work. Microsoft and AWS are quietly conceding that software alone can't close the gap between what AI does in a demo and what it does at your company, so they're spending billions to hire their way across it. The winners of the next phase of AI are probably going to be the companies with the most engineers willing to show up in person and stay until something actually ships. Benchmark leaderboards are becoming a sideshow to what happens on the customer's laptop at 4 p.m. on a Tuesday.
