On Thursday, Judson Althoff, the CEO of Microsoft’s commercial business, announced Microsoft Frontier Company. Microsoft is investing $2.5 billion in Microsoft Frontier Company, a new AI business dedicated to embedding engineers directly into client organizations.
The CEO disclosed that Microsoft is investing $2.5 billion in the unit and will staff it with 6,000 “industry and engineering experts.” The initiative is Microsoft’s move to capture the market, using existing AI models to do something useful within a Fortune 500 company, rather than building new, bigger models.

According to Althoff, Microsoft Frontier Company will be physically embedded with clients to support design, innovation, and deployment, and to improve AI systems at scale continuously. Microsoft is essentially building a team of experts to help AI consumers understand how best to use it, rather than leaving customers to figure it out on their own.
Judson Althoff clarified that this program is far bigger than what the industry now refers to as “Forward Deployed Engineer (FDE).” In his statement, he mentioned that the Microsoft Frontier Company goes beyond FDEs. However, Althoff failed to clearly distinguish what actually makes it different from other FDEs that Palantir and AWS already run.
In the past few weeks, major tech players in the AI space have been rushing to adopt the FDE model. Just recently, Amazon unveiled its own version, a $1 billion push led by VP Francessca Vasquez. Amazon’s model seeks to integrate engineers in pods of five or six for roughly 45-day sprints with clients like the NFL, NBA, and Southwest Airlines.
Earlier in the year, OpenAI stood up its own deployment company backed by outside investors, including TPG and Bain Capital, at a reported $4 billion valuation. Anthropic also built a similar vehicle backed by Blackstone, Hellman & Friedman, and Goldman Sachs at around $1.5 billion.
The only major difference here is that Microsoft’s $2.5 billion is entirely self-funded, with no outside capital. The self-funded approach is more similar to AWS’s than the venture-backed structures at OpenAI and Anthropic.
In its blog post, Microsoft points to work already underway with LSEG to embed AI into its Workspace product for finance professionals, as well as engagements with Land O’Lakes, Unilever, and Novo Nordisk.
Microsoft’s share price has dropped roughly 20% this year, by far the worst performance among mega-cap tech companies. Copilot adoption hasn’t reached the scale investors hoped for, and GitHub Copilot has been losing ground to newer coding agents. Many analysts argue that the underlying fear for most investors is AI’s ability to do what Microsoft has been doing for years, building software.
Microsoft stock responded to the news, rising steadily after the opening bell, gaining over 2% within the first few hours. At the time of publication, MSFT is trading above $390, with a strong buy recommendation from over 36 analysts in the past 3 months.
Althoff and his team are betting that by embedding AI engineers inside client companies, they can turn AI pilots into AI that actually sticks. And according to Microsoft, this is a $2.5 billion opportunity.
The smartest crypto minds already read our newsletter. Want in? Join them.


