Microsoft stands up $2.5B 'Frontier Company' to embed 6,000 engineers inside AI customers

The venture, led by former Microsoft Asia president Rodrigo Kede Lima, is the largest entry yet in the forward-deployed AI engineering race.

Microsoft unveiled a $2.5 billion operating unit on Thursday called Microsoft Frontier Company, staffed with 6,000 engineers and industry specialists who will embed inside enterprise customers to build and run their AI systems. Rodrigo Kede Lima, most recently president of Microsoft Asia, will lead it. The pitch is co-designed deployments with measurable business outcomes, protection of customer data and IP, and customer ownership of the systems produced, with support for both Microsoft’s models and third-party ones against each buyer’s internal data.

Judson Althoff, CEO of Microsoft’s commercial business, framed the scale bluntly. “This goes beyond what has been labeled as Forward Deployed Engineering, and will be the largest, most capable, outcome-driven engineering organization in the industry,” he said. The reference point is Palantir, whose forward-deployed engineers turned services-inside-the-customer into a defensible enterprise motion, and which TechCrunch notes is running similar engagements on Nvidia open-source models.

The timing is the interesting part. AWS committed $1 billion to its own embedded-engineer initiative two days earlier. OpenAI and Anthropic have launched comparable deployment groups this year. Every hyperscaler and frontier lab has now concluded that the binding constraint on AI revenue isn’t model quality but integration labor, and that whoever ships the engineers ships the account.

Althoff was candid about why Microsoft is moving now. “Three years ago, when we built Copilot, we made a mistake by binding it to OpenAI models only,” he told Reuters, citing gains by DeepSeek and Google’s Gemini. Frontier Company’s model-agnostic posture is the correction.

Some details remain deliberately unclear. GeekWire notes that Frontier Company isn’t a separate legal entity, that most of its 6,000 people already work at Microsoft alongside salespeople being folded in, and that Microsoft hasn’t disclosed whether the $2.5 billion is fresh spending or repurposed budget, or over what period. The number is a narrative artifact as much as a capital commitment, and it’s meant to be read against the AWS billion. In enterprise AI, the services layer is the moat now, and the incumbents are pricing it accordingly.