Microsoft Corp. has formed a new business that will help organizations build and manage artificial intelligence applications.

The Microsoft Frontier Company, as the venture is called, launched today with an initial $2.5 billion investment from the tech giant. It’s staffed by 6,000 “industry and engineering experts.” The group is led by Microsoft Corp. executive Rodrigo Kede Lima, who was previously the president of Microsoft Asia.

The unit will speed up enterprise customers’ AI projects by sending forward-deployed engineers, or FDEs, to their offices. The FDEs will help clients’ in-house development teams build custom AI applications. They will use FinOps, a set of technology-focused financial analysis methods, to determine their projects’ return on investment.

The Microsoft Frontier Company’s engineers will also help customers “continuously improve AI systems.” Some software teams fine-tune their large language models on a regular basis to address changes in user request patterns. Without such adjustments, an LLM can experience a decrease in output quality.

The unit presumably plans to use the tech giant’s own cloud services to power the AI software that it will build for customers. Azure includes a suite of services called Microsoft Foundry that can speed up AI development projects. It offers safety guardrails, model training tools and related building blocks. 

Microsoft Foundry also provides access to more than 11,000 hosted AI models. The library features both third-party algorithms from partners such as Anthropic PBC and internally developed LLMs. The lineup includes MAI-Thinking-1, Microsoft’s newly introduced flagship reasoning model. The company says that the LLM is competitive with Claude Opus 4.6 on the SWE-Bench Pro coding benchmark and uses significantly less hardware.

The Microsoft Frontier Company will team up with other professional services providers to support international customers. The business’ initial partner roster includes Accenture plc, Capgemini SE, EY and other major industry players.

“ It will provide a unique combination of skills inclusive of deep industry knowledge, change management and continuous improvement experience, and enterprise-grade AI engineering expertise,” Judson Althoff, chief executive of Microsoft’s commercial business, wrote in a blog post.

The company’s rivals are also enhancing their forward-deployed engineering capabilities. On Tuesday, Amazon Web Services Inc. announced plans to invest $1 billion in a dedicated FDE organization that will help customers build AI agents. Google LLC’s cloud business, in turn, recently launched an FDE recruiting initiate.

Anthropic and OpenAI Group PBC have also formed FDE organizations this year, but they’re taking a different approach from the major cloud providers. Both companies are partnering with external investors to bring their forward-deployed engineering services to market. Anthropic reportedly raised $1.5 billion for its FDE venture while OpenAI has secured $4 billion from a SoftBank Group Corp.-backed consortium. 

Photo: Microsoft