Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Network resilience as the backbone

This key concept supports both AI and digital infrastructure.

[Photo: Getty Images]

Profile photo of Ashley Sturm

VP of Global Marketing · Opengear

There is a lot of energy around AI right now. Real investment and ambition have driven deployment across industries faster than most enterprises anticipated, even two years ago.

From predictive analytics in financial services to autonomous systems in manufacturing, AI is becoming embedded in how modern organizations operate. The conversation has shifted from “should we” to “how fast can we,” and that momentum is not slowing.

Yet the infrastructure conversation hasn’t kept pace, especially at the network layer, which links an organization’s computers, data centers, cloud environments, and applications so they can communicate and operate.

The network layer gets less strategic attention than compute, storage, and software. Yet when the network becomes an afterthought, it can limit the return on everything else.

The architectural decision

Most enterprises are already making an architectural decision about resilience—their ability to keep critical systems running and recover quickly when something goes wrong.

For example, imagine AI workloads spanning multiple data centers or cloud environments. Without a separate way to access and manage the underlying infrastructure, IT teams can lose the ability to troubleshoot operations if the primary network goes down. Operational infrastructure control relies on a network that may have failed.

Most leadership teams wouldn’t consciously sign off on that, but resilience has historically lived in infrastructure planning rather than strategy. An encouraging shift is underway as more enterprises focus on the question, “How do we ensure we always have control?” This leads to a stronger position.

AI moves closer to the edge

Enterprise AI’s centralized first wave included large data centers, public cloud platforms, and controlled environments with dedicated technical staff nearby.

The next wave looks different, as AI moves into retail stores to analyze real-time customer behavior. Hospitals rely on connected diagnostics, and industrial operations run AI-driven monitoring across geographically dispersed facilities.

Enterprises now manage hundreds or thousands of sites, not a handful. Branch offices, edge locations, data centers, and cloud environments work together to support modern workloads. The new scale requires effective infrastructure management, including remote access. Remote access is a core operational requirement when running AI workloads across healthcare, finance, transportation, energy, and other sectors.

Our principal architects at Opengear regularly work through reference architectures with enterprise teams. Reference architectures are proven infrastructure blueprints that help organizations plan how to build and connect their technology systems.

One consistent, straightforward finding: Environments that include a separate management connection let IT teams monitor, access, and troubleshoot critical infrastructure even when the primary network is down.

For example, if a network outage prevents employees from accessing business applications, IT teams can still use this independent connection to identify the problem and restore service. As a result, teams diagnose faster, restore operations faster, and operate with more confidence as they scale.

3 components of resilient infrastructure

Resilient infrastructure comes down to three things working together as a system: intent, insight, and control.

Intent defines the baseline

Intent means knowing what the network was designed for. Every infrastructure environment is built to support specific workloads, connect specific systems, and behave in defined ways under defined conditions.

Intent states that design clearly. Enterprises that build with intent have defined “normal,” which becomes the standard against which everything else is measured.

Insight reveals reality

Insight means having a continuous, up-to-date picture of how infrastructure is actually performing compared with how it is supposed to operate. Resilient architectures keep operational insight independent of the primary network, so the people responsible for a fix can still troubleshoot.

Control is what you do with reality

Control is what out-of-band management is built to provide. When observed state does not match intended state, someone has to be able to act. The ability to reach infrastructure when the primary network is down is foundational for large enterprises.

Without it, recovery from a remote site outage may require dispatching someone physically to the location. For enterprises with global operations, that recovery path can take hours or days. For enterprises running AI-driven automation, those delays slow operations, stall data processing pipelines, and cause disruptions to cascade across interconnected systems. Control must be architected in. It does not appear by default.

Resilience as a competitive position

Enterprises with resilient infrastructure operate differently. They move into new environments faster and scale distributed operations, without the exposure that comes from unavailable infrastructure.

Performance and scalability are necessary, but no longer sufficient. Enterprises that pull ahead with infrastructure keep it manageable under pressure, because that consistency increasingly separates operational leaders from everyone else.

Gartner and IDC have both pointed to network resilience and out-of-band management as growing enterprise infrastructure priorities, particularly as edge deployments scale, and AI workloads carry more operational weight. The market is arriving at conclusions that leading enterprises reached some time ago.

Resilience built in from the start costs less, performs better, and creates more strategic flexibility than retrofitting after an incident makes the gap undeniable.

The foundation AI depends on

The rapid growth of AI, edge computing, and connected systems is reshaping the technology landscape in ways worth optimism. The opportunity is real. Capturing it fully, though, depends on infrastructure remaining stable, secure, and consistently within reach.

Network resilience makes that possible. It enables enterprises to manage infrastructure remotely, maintain security controls during incidents, and restore services without waiting for physical intervention. For any enterprise serious about the long-term return on its AI investment, resilience belongs in the architecture conversation from day one.

AI may be the defining technology of this period, but its success depends on something more foundational. Intelligent systems deliver value when the supporting infrastructure stays reachable, manageable, and built to last.

Ashley Sturm is vice president of global marketing at Opengear.