Agentic transformation starts with leadership, not technology:
The operating model debt that AI agents have exposed
Mission-critical operators don’t have an agent problem. They have a decades-old structural decision that agents just made impossible to ignore.
When you spend your week on calls with organizations figuring out how to deploy agents successfully, you start seeing patterns. Glenn D’Haene sees them at Skyline Communications, working with operators across telecom, satellite, and media.
“Every conversation reveals the same challenges, but the actual problem is that leadership keeps misdiagnosing them. They think the problem lies with the new emerging technology. It doesn’t. The barrier for change is the organizational operating model. And they’re running out of time to fix it.”
The barrier for change is the organizational operating model. And they’re running out of time to fix it.
What agents are actually exposing
Most of the conversations around agents are framed as a technology challenge: Which model to use. How to scale pilots. Where to find the talent. How to handle security. But that framing, while instrumentally important, largely misses what is actually happening.
“Usually, it’s not an agentic problem at all. It’s the accumulated debt of two decades of decisions,” explains Glenn. “Operators have years and years of systems being siloed or connected via APIs, Kafka, and service buses. An architecture built for connecting data between departments and people, but it never created a unified fabric.”
Usually, it’s not an agentic problem at all. It’s the accumulated debt of two decades of decisions,
And that unified fabric is essentially what agents need to thrive. Because it unifies and contextualizes real-time and historical data, real-time controls, and machine-learning KPIs across your entire organization in one foundational model that agents can reason over and act on. In practice, this requires a purpose-designed platform foundation.
“And that’s crucial. Agents are quickly becoming true craftsmen. They’re exceptional at taking action at scale and speed. But they reason only as well as the granularity and context of the data and controls they’re given. And that’s where the real challenge lies. Because most organizations have a connected patchwork instead of a unified fabric.”
For many years, that connected patchwork of systems, spreadsheets, and email threads worked just fine. But as companies have started experimenting with agents at scale, the seams start to show.
The handover nobody planned
To understand why the foundation is so scattered, you have to go back to how it happened.
The digital transformation from hardware to software was genuinely complex. Someone had to manage it, and all eyes were on IT because they had the skills and the tooling. They were handed the keys and operations teams had to take a step back.
“But IT and operations are different jobs. One optimizes for system stability, the other for business impact. And both departments have their own budget, ruling, and management. So you end up with islands optimizing for completely different priorities, and essentially competing against each other for the same company resources. It drains energy and kills innovation.”
Another issue Glenn flags with treating transformation as an IT job: IT mastered building data platforms to analyze the world, but never prioritized real-time control and orchestration.
“Great insights don’t automatically deliver faster actions. And today’s revenue engine simply moved from getting insights to making real-time decisions. If your control capabilities are not part of your unified data stack, decisions will always be a bottleneck, whether people make them or agents.”
If your control capabilities are not part of your unified data stack, decisions will always be a bottleneck, whether people make them or agents.
The friction between departments is not a personality conflict, but the structural cost of an operational model that was never designed for an agentic future. And it will not be resolved at the department level.
“It’s really leadership that needs to reset. You have to look at your organizational operating model and say: this doesn’t fit anymore. And it’s not just about technology — it’s about people, departments, budgets, how procurement operates. It’s really a strategic shift in how you organize your company tomorrow.”
Reclaiming the operating model
So what does that strategic shift actually look like? It’s definitely not abstract reorganization. It requires concrete structural changes, says Glenn:
Restructure how operations are budgeted
Stop funding tools at the department level and move toward cross-functional platform ownership. When departments each buy their own monitoring stack, you get islands. When you fund a unified platform that serves all of them, accountability changes. So does speed. The question isn’t “which tool should we buy?” but “who owns the operational foundation everyone depends on?” For satellite operators, this is often where SatOps decisions start.
Elevate satellite services with ICT excellence
Deliver broadband, voice, data, media, IoT, and EO services across any satellite and telco network commercial space, government, and defence.
Clarify accountability for the operational data layer
Someone needs to own the unified data and control plane, not as an IT project, but as a transformational asset. And critically, that decision must be driven by business logic, not IT or engineering convenience. IT’s job is to build and scale data, analytics, and back-office systems; operations’ job is to enable real-time operational insights and decisions on a service orchestration level, and build autonomous networks and next-gen operational teams powered by agents. Get that relationship backwards and you’re back to analytics-first thinking.
Reframe agentic readiness as a management question, not a technology one
Don’t ask “which AI model should we deploy?” Ask instead: “Are we structured to actually use agents? Do we have unified data? Do we have real-time control? Do we have cross-functional teams who can act on what agents tell us? Have we organized our procurement and budgets to act on it?” If the answer to any of those is no, the technology won’t save you.
Recognize that your revenue engine is shifting
For years, competitive advantage came from better data insights. Today, it’s faster action. An organization with perfect visibility into network anomalies but 4-hour remediation cycles loses to one with less visibility but 1-minute autonomous fixes. Operations isn’t a cost center to optimize anymore. It’s where differentiation happens, and should be treated as your next-generation revenue engine. Budget accordingly, especially in environments like MediaOps where execution speed directly affects service outcomes.
Harmonizing ICT and media for better planning, live operations, and file media operations
Transform your operations by creating a better, faster, and more cost-effective environment that is enjoyable to work in!
The speed vs. perfection trap
Understanding the restructuring is one thing, executing it is another. Even leaders who grasp what needs to change get caught on two pitfalls.
This is not the moment to build from scratch
“The operators moving fastest right now are not building infrastructure. They are making a platform decision, deploying fast, and immediately redirecting their teams toward the work that actually differentiates them: the products, the services, and the customer experience that no platform can build for them.” The commercial model behind that platform decision also matters, including how teams scale responsibly over time.
Go-to-market cycles are way shorter than ever. The cycles that used to take five years now take one. If you still need to start building your unified data stack properly now, you’re already quite late. You simply can’t afford to spend two to three years constructing a foundation while competitors already running on one are pulling further ahead.
Don’t let governance paralyze your timeline
Sovereignty, data residency, and compliance all matter, but they shouldn’t paralyze your timeline. Tackle them wisely as you scale, but do it on a vetted, proven platform deployed now, rather than building perfect governance before you start.
The operators moving fastest aren’t waiting for zero-risk approval; they’re deploying on proven platforms and adapting their governance as they scale, rather than treating perfect compliance as a prerequisite to launch.
Leaders, it’s time to act
The bill for twenty years of fragmented operational decisions is due. Agentic didn’t create it, but it certainly amplified it.
The organizations that act, that take back control, unify it, and stop waiting for the technology to somehow fix an organizational problem, are the ones that will still be setting the pace in three years.
The ones that don’t will be explaining to their boards why the leaner competitor just took another three accounts.
I think there are a lot of leaders out there right now thinking: we need to switch gears. To them I say, do it.
“I think there are a lot of leaders out there right now thinking: we need to switch gears. To them I say, do it. This is a leadership decision disguised as a technology decision. Take back control and shift now before it’s too late.”