Microsoft creates “Microsoft Frontier Company,” a $2.5 billion unit to embed AI engineers at enterprise customers

Microsoft is launching a new enterprise-focused organization called Microsoft Frontier Company, committing $2.5 billion and planning to embed thousands of engineers directly inside customer operations to accelerate the deployment of artificial intelligence systems, according to a company announcement posted July 2.

Microsoft said Frontier Company will combine deep industry knowledge, change management and continuous improvement experience, and “enterprise-grade AI engineering expertise.” The company described the effort as going beyond a work model it refers to as Forward Deployed Engineering, positioning it instead as its largest and most capable outcome-driven engineering organization in the industry.

The company said Frontier Company will help customers deliver “Frontier Transformation” through AI by co-designing, co-innovating, and deploying solutions alongside client teams. In the announcement, Microsoft tied the push to a shift in enterprise demand from early-stage AI trials toward the ability to prove business results and returns on AI spending.

Microsoft framed the challenge as two-part: getting “intelligence” to compound over time while maintaining “trust.” It said enterprises need an “intelligence platform” that grows from a company’s proprietary data, expertise, workflows, and decision-making processes, using their choice of models to build AI solutions and routines.

The company also highlighted the need for a trusted platform to observe, govern, manage, and secure AI solutions across different layers of the technology stack. Microsoft referenced cost and ROI assessment using FinOps, a set of practices used to manage cloud and data costs and measure financial performance, and said customers can use “different levers” to manage cost.

Beyond engineering, Microsoft said Frontier Company will be staffed with 6,000 industry and engineering experts and will be designed to place teams with customers as they adopt AI. That model targets a practical bottleneck enterprises face, where teams can pilot AI but struggle with deployment at scale and with integrating AI into existing operations and controls.

Separately, reporting from New Orleans CityBusiness said the initiative is also aimed at helping businesses combine AI models from Microsoft and third parties while retaining ownership of the resulting work, and noted that customers want flexibility to switch among AI models rather than rely on a single provider.

Still, Microsoft did not lay out detailed commercial terms in the announcement, such as pricing, customer eligibility, or whether Frontier Company will operate as a separate subsidiary with standalone contracts. It also did not specify the timeline for reaching full staffing or which industries or regions will be prioritized first.