For decades, system integrators have been the bridge between enterprise technology and its actual implementation. But the rise of generative AI and the pressure to demonstrate return on investment are reshaping that role. Microsoft and Amazon Web Services (AWS) have announced multi-billion-dollar investments to deploy their own engineers directly into client companies, a strategy that promises to accelerate AI adoption and close the gap between experimentation and production.

The Context: The Investment-ROI Gap Widens

According to Thomas Randall, research director at Info-Tech Research Group, organizations are under increasing pressure to demonstrate the productive value of their AI deployments. "The gap between AI investment and return on investment continues to grow," he says. In this scenario, so-called Forward Deployed Engineers (FDEs) emerge as a solution to shorten learning curves, establish reusable processes, and transfer capabilities to internal teams.

Microsoft has introduced Microsoft Frontier Company, a $2.5 billion initiative that will embed 6,000 experts alongside clients to co-design, co-innovate, and implement AI systems aligned with their business objectives. Meanwhile, AWS has announced a $1 billion investment in its AWS FDE platform, which also incorporates experienced engineers into client teams.

Beyond Traditional Consulting

Judson Althoff, CEO of Microsoft Commercial Business, explains that the offering focuses on what they call Frontier Transformation: helping clients build an intelligence platform based on their proprietary data, workflows, and decision-making processes. The proposal is grounded in FinOps principles, allowing users to observe, govern, manage, and protect their AI tools across the entire infrastructure. "Clients should not be locked into a single model just as they should not be locked into a single technology provider," Althoff emphasizes, highlighting that the platform is open and heterogeneous, compatible with ChatGPT, Claude, Microsoft Copilot, and other models.

Francesca Vasquez, vice president of engineering and frontier AI services at AWS, agrees that their approach goes beyond traditional consulting. "Unlike traditional consulting, which assesses, recommends, and treats each deployment as an independent project, AWS FDE builds with the long term in mind," she says. Clients transition from observers to co-developers and ultimately to autonomous operators, acquiring competencies and reusable patterns.

Use Cases and Measurable Results

Early users of Microsoft Frontier Company, such as London Stock Exchange Group (LSEG), Land O’Lakes, Unilever, and Novo Nordisk, are already seeing tangible results. For example, AI integrated into LSEG Workspace allows financial experts to ask complex questions and get quick answers based on structured and unstructured data. The technology base is iteratively refined through real-time feedback and testing, accelerating each cycle and improving model quality.

In AWS's case, Vasquez highlights that the platform is designed with an agentic-first strategy, aimed at reducing deployment timelines from months to days. Embedded engineers, many of whom develop AWS's own AI services, validate and guide projects. Additionally, clients gain access to operational guides (runbooks) and architectural documentation, while a semantic layer connects to their data sources to create a knowledge graph on which AI agents can reason.

A critical aspect is the preservation of institutional knowledge. Vasquez notes that specialized domain knowledge is captured in code, agents, and systems, preventing loss due to staff turnover. Security tools provide hardware-based isolation and end-to-end encryption.

The Role of System Integrators

System integrators have enjoyed high-margin relationships with their clients for decades. Carmi Levy, a technology analyst, believes it "makes perfect sense" for hyperscalers to try to capture some of that business. "Both Microsoft and Amazon are aggressively seeking new ways to reinforce client dependency and open more opportunities to influence their operations," he says.

However, Randall notes that FDEs and integrators offer complementary services. While FDEs focus on AI operating systems, reference architectures, and runbooks, integrators bring broader knowledge of cross-system integration, change management, and program scaling. "Their deliverables will be more strategic and far-reaching," Randall says. Microsoft has already announced it will work closely with global partners such as Accenture, Capgemini, EY, KPMG, and PwC to scale its platform.

Implications for Businesses: What Should They Consider?

For IT leaders, these new options offer an accelerated path to AI production, but they are not without risks. Levy recommends analyzing not only technical capabilities but also whether vendor motivations align with client interests. "Assuming Microsoft and Amazon offer competitive pricing and service levels, they could represent a very attractive alternative, but relying on their services could limit long-term choice," he warns.

Randall advises companies to reflect on the type of outcome they seek. FDEs are more suitable for organizations that want to move beyond pilots and quickly advance to effective, operational products. System integrators become necessary when scaling that model across complex, heterogeneous business processes. "FDEs are not suitable for organizations still resolving basic AI strategy issues or that wish to maintain a cloud-neutral stance," he concludes.

In a context where 77% of organizations lack a corporate-level AI strategy, according to Info-Tech, the emergence of hyperscaler FDEs can be a catalyst or a dependency trap. The final decision will depend on each company's technological maturity and appetite for strategic risk.

To delve deeper into protecting the infrastructures that support these deployments, we recommend our guide on setting up secure VPNs and firewalls, as well as the analysis of the triple alliance against exploits in the AI era. If your team is adopting AI agents, don't miss the reflection on quality and mentoring arising from Godot's decision to ban AI agents.

Original source: ComputerWorld. Analysis and adaptation by ForgeNEX.