The race to deploy agentic AI in the enterprise has shifted from software licensing to human capital. Microsoft and Amazon Web Services (AWS) are committing billions of dollars to embed thousands of engineers directly into customer organizations. This move targets the gap between AI investment and return on investment, aiming to compress deployment timelines and transfer skills internally.
Microsoft’s $2.5 billion Frontier Company
Microsoft has launched Microsoft Frontier Company, a new venture backed by a $2.5 billion investment. The initiative deploys 6,000 experts to work alongside customers to co-design, deploy, and improve AI systems. Judson Althoff, CEO of Microsoft Commercial Business, describes this as “Frontier Transformation,” focusing on building intelligence platforms based on proprietary data and internal workflows.
Unlike traditional consulting, this service emphasizes long-term capability building. The platform is described as model-diverse and open, allowing customers to choose from ChatGPT, Claude, Microsoft Copilot, or other open-source models without vendor lock-in. Microsoft asserts that customer data and intellectual property are protected and not used to train its own models.
Microsoft will leverage partnerships with systems integrators like Accenture, Capgemini, EY, KPMG, and PwC to scale the platform. Early adopters including London Stock Exchange Group (LSEG), Land O’Lakes, Unilever, and Novo Nordisk are already using the service. For example, LSEG uses embedded AI in its Workspace tool to answer complex financial questions based on structured and unstructured data.
AWS’s $1 billion FDE platform
Amazon Web Services announced a parallel commitment with a $1 billion investment into its own AWS Forward Deployed Engineer (FDE) platform. Francessca Vasquez, VP of frontier AI engineering and services at AWS, explained that these engineers embed into business, engineering, and security teams to build agents tailored to specific data and governance frameworks.
The goal is to move customers from “observers to co-builders to autonomous operators.” AWS claims this agentic-first approach can compress project timelines from months to days. The service provides runbooks, architectural documentation, and a semantic layer that connects to data sources to create knowledge graphs for AI agents.
AWS emphasizes that this service is not for experimentation but for organizations ready to run production AI systems. Security measures include hardware-based isolation and end-to-end encryption. The platform aims to ensure institutional knowledge remains within the customer’s code and systems, preventing loss due to employee turnover.
What this means for you
For enterprise IT leaders, these initiatives offer a potential shortcut past the AI strategy gap. Research indicates 77% of organizations lack a corporate-wide AI strategy. FDEs provide narrow, specific expertise to fast-track builds on specialized platforms. However, analysts note that traditional systems integrators (SIs) still hold value for broader integration across messy enterprise processes and change management.
Decision-makers should evaluate whether they need quick, effective product buildouts via FDEs or broader strategic scaling via SIs. Considerations include vendor lock-in risks and the specific competencies of the embedded engineers versus existing partners.
Over to you: Would your organization trust an embedded vendor engineer more than a traditional systems integrator for critical AI projects?
You may also like
