Every business is under pressure to do something with AI. Boards are asking questions, vendors are making promises and there is a mounting sense of urgency to innovate or fall behind. But the conversation most businesses are not having is about what sits underneath all of that.
“AI introduces more and more unknowns into how things work,” says Neil Dragt, Echo Innovation Executive. “You can't assume AI will magically solve problems, you've got to have a platform you understand and trust, so you can see what's going on.”
Simply put, AI does not replace the need for stable, well-managed observable networks, it intensifies it. “AI almost adds a chaos factor. It's non-deterministic, it never responds the same way twice,” he continues. “Every app you open, whether it's Word or Gmail, has a little chaos monkey in the corner making things unpredictable.” The result is a network environment that is harder to predict, troubleshoot and manage – and most businesses do not have the visibility to deal with it.
“For me, visibility is the network tools that give you insights into what's happening,” says Dragt. “In the world of AI, the need for those insights hasn't gone down, it's gone up.”
In practice, visibility operates across three layers. At the network level, it means understanding what traffic is flowing, where it is going and whether it is behaving as expected. At the application level, it means knowing which applications are running, how they are performing and whether they can be steered and prioritised. At the device and security level, it means knowing what is on the network and what it is doing, and underpinning all three is real-time data, because data that is an hour old, says Dragt, is not truly useful visibility. “If you centralise your network data in real-time, you can start answering all the important questions,” says Dragt. When something goes wrong (and in complex environments, something always does) the ability to diagnose it depends entirely on visibility.
Seeing is knowing
Most businesses underinvest in visibility and Dragt says that this usually starts with a purchasing decision. Take a mid-market business replacing a firewall, for example. If there are managed, observable options and a cheaper unmanaged alternative, most would choose the cheaper option. “If you can't see what's happening, you can't fix it and you can't control it,” says Dragt.
In his experience, what the cheaper option does not provide is application-level insight, security integration or any meaningful forensic trail. There is no way to steer traffic, no way to understand what individual applications are doing and no evidence base when something goes wrong. For a business running a handful of standard applications on a stable network, that might be manageable, but for one running AI-enhanced SaaS tools, cloud services and mobile workforces, it is a structural problem.
AI changes the network environment in ways that make good foundations imperative. Because AI-driven applications are variable by nature, traffic patterns shift constantly and there is no stable profile to plan against, which means monitoring and insight has to be richer and more real-time than ever before.
“AI helps you make sense of all the real-time data these technologies collect, and it can make visibility more efficient. But it's also creating a lot of the problems,” says Dragt. AI-powered threat detection can surface issues and filter out noise, but only if the underlying data is there and a network without centralised real-time telemetry gives those tools nothing to work with.
Combat complexity
A managed SD-WAN and SD-Branch environment combined with a managed firewall gives businesses centralised visibility across devices, applications and network flows. Echo builds this on Fortinet's FortiGate technology and Security Fabric, with FortiAnalyzer providing the analytics layer that turns raw telemetry into something actionable.
“It's as important for smaller organisations as it is for the banks,” adds Dragt. “You still need the right enabling components; we just help you manage that complexity.”
For mid-market customers, the outcome is enterprise-grade visibility and control without requiring internal expertise. Echo manages the stack, which means the data being collected is also being interpreted and acted on, resulting in faster diagnosis, more predictable performance and a clearer evidence base for compliance. “Invest a little more in your tools, let us help you manage them and you'll be prepared for the complexity that's coming,” says Dragt.
AI may be changing the network environment, but you still can't forget the fundamental building blocks of modern connectivity: visibility, determinism and the right infrastructure to support both. And the businesses that invest in managed, observable networks can introduce new AI‑driven tools without sacrificing control. “Don’t just adopt the fancy new AI because someone on your board told you to. Make sure you can monitor it, keep it working and run it securely first,” ends Dragt.
Share
