Microsoft’s announcement follows similar moves across the AI industry. Earlier this week, Amazon said it would invest $1 billion in its own forward deployed engineering initiative to support AI projects. Anthropic and OpenAI also established FDE groups in May, partnering with private equity firms, banks and consulting firms.
The expansion comes as Microsoft continues to invest heavily in its AI strategy. The company has spent tens of billions of dollars building data centers to power generative AI models while rolling out a range of AI products. Adoption has varied across its portfolio, with Microsoft 365 Copilot yet to become widespread in business settings and GitHub Copilot facing stronger competition from newer coding assistants.
According to Judson Althoff, chief executive of Microsoft’s commercial business, the company created the new division after recognizing that customers remain at different stages of AI adoption and are still working through key implementation decisions. “Customers are in very different places right now, and trying to really figure out AI,” Althoff said in an interview.
He said many organizations are still deciding fundamental questions about their AI strategies. “Do they snap to one model from OpenAI or one model from Anthropic, or a family of models?” Althoff said. “Do they take it from a technology first mindset? How do they look at their existing business processes and operations?”
Althoff said Palantir helped popularize the forward deployed engineering role. He noted that Microsoft’s approach differs by offering support for a broader range of AI models, data connectors and integrations. “More models, we support more connectors to data, more integrations with open systems of record,” Althoff said.
Microsoft has long offered implementation and support services to enterprise customers. In the March quarter, the company generated about $2.1 billion in revenue from enterprise and partner services, an increase of 2.5% from a year earlier.
Althoff said Microsoft’s strongest customer outcomes have come from taking a structured approach to AI deployments. The company has had the most success when it takes a “very methodical approach towards working with customers to build out an intelligence platform” that protects their intellectual property and allows them to take advantage of “any model in the ecosystem,” he said.
This analysis is based on reporting from CNBC.
Image courtesy of agolution.
This article was generated with AI assistance and reviewed for accuracy and quality.
