Microsoft introduced Microsoft Frontier Company on July 2, a new operating business built to embed engineering and industry experts directly inside customer organizations to design, deploy, and continuously tune agentic AI systems. Commercial Business CEO Judson Althoff announced the unit, which is led by Rodrigo Kede Lima, backing it with a $2.5 billion investment and roughly 6,000 industry and engineering specialists.

What Frontier Company does

Rather than selling software licenses alone, Frontier Company places Microsoft staff inside client organizations to co-design and operate AI systems built on customers' own proprietary data, workflows, and choice of underlying models. Microsoft describes the offering as going further than typical "forward deployed engineering" arrangements, pairing deep industry expertise and change-management support with enterprise-grade AI engineering to help agentic processes keep improving over time. Customers also get a platform to observe, govern, and secure the AI systems Frontier Company helps build, addressing a recurring enterprise concern about handing agents access to sensitive internal data and processes.

Part of a broader industry pattern

The launch lands two days after Amazon committed $1 billion to a comparable initiative, and follows similar forward-deployment ventures that OpenAI and Anthropic launched in May. The convergence suggests major AI vendors increasingly see hands-on deployment services, not just model access, as the bottleneck standing between enterprises and production AI adoption.

Why it matters

Enterprise AI rollouts have repeatedly stalled on integration with legacy systems, data governance, and change management rather than on model capability alone. By putting its own capital and headcount behind direct deployment work, Microsoft is betting that services-led AI adoption, not self-serve tooling, will determine which vendor captures the largest share of enterprise AI spending as agentic systems move from pilots into core business processes.