Microsoft is making a deeper push into enterprise AI services with the launch of Microsoft Frontier Company, an organization backed by a $2.5 billion investment that will help customers deploy and manage AI systems customized to their business.
The new organization brings together more than 6,000 engineers, technical consultants, sales specialists and industry experts under leadership headed by Rodrigo Kede Lima, formerly president of Microsoft Asia.
The debut of Frontier Company addresses a challenge facing many enterprise customers. Rather than relying on a single provider, many organizations are combining models from OpenAI, Anthropic, Google and open-source developers. Integrating those technologies with corporate data and existing workflows is far more difficult than many businesses anticipated.
The Rise of Forward Deployed Engineering
Microsoft’s approach relies on forward-deployed engineering, in which technical teams work inside customer organizations to build and operate AI systems. Instead of delivering software alone, engineers collaborate with customers to identify practical use cases and oversee implementation.
Forward deployed engineering has become a competitive segment of the AI market. AWS recently committed $1 billion to a similar initiative, while OpenAI and Anthropic both launched organizations this year focused on embedding engineers with enterprise customers. Microsoft is positioning Frontier Company as a full-featured effort that combines engineering expertise with industry-specific consulting.
The strategy’s focus on multiple AI models marks a change for Microsoft’s enterprise AI strategy. The company initially tied Copilot exclusively to OpenAI models, but as competing models like Google’s Gemini and China’s DeepSeek matured, enterprise customers wanted to select different models for different workloads. Microsoft now argues that business value comes from combining proprietary enterprise data with whichever AI model best fits a particular task, rather than committing to a single provider.
This “use the best model” approach acknowledges that AI models are becoming commodified, and that implementation services are a better revenue source than focusing on a single model.
As with other forward deployed engineering initiatives, Microsoft stresses protecting customer intellectual property. The company says organizations using Frontier Company will retain ownership of the data and operational knowledge developed through these engagements rather than contributing those assets to AI model training.
Microsoft also already operates consulting and implementation organizations, making Frontier Company an expansion of existing capabilities rather than an entirely new business. Still, combining thousands of specialists into a dedicated AI-focused operation supports Microsoft’s expectation that implementation services will become a competitive advantage as AI models gain still more adoption.