Tony Vrushaj - Episode 169 - The Route to Networking

From Curiosity to Orbit: Tony Vrushaj on Data Centres, AI, and Why the Future Might Be in Space

A conversation between Jamie Maher and Tony Vrushaj on the Route to Networking podcast

Data centres were once invisible infrastructure, humming quietly behind the internet.

Today, they are geopolitical assets, AI engines, and energy-hungry powerhouses shaping the global technology race.

In the age of AI, cloud, and high-performance computing, data centres are no longer background infrastructure. They are strategy.

In this episode of The Route to Networking podcast, Jamie Maher sits down with Tony Vrushaj, Data Centre Team Leader at IBM, to explore how the industry has evolved over the last decade, what AI is changing beneath the surface, and what the next generation must understand if they want to build a serious career in data centre infrastructure.

Tony’s story is not a straight line. It is built on curiosity, persistence, mindset, and a willingness to step into environments where failure is expensive and growth is inevitable.

A Non-Linear Route into Data Centres

Tony’s career did not begin in a server room.

It began under the hood of a tractor.

“It was curiosity… I don’t really have a linear career,” he explains. “It started with helping my dad fixing his tractor when I was five.”

From tractors to cars, then to fixing computers for friends and family, his path into tech was driven by a simple instinct to understand how things work.

After moving to the UK in 2014, Tony worked in hospitality, demolition, and at a train station while improving his English. But the pull toward technology never disappeared.

While working shifts, he enrolled in courses including CCNA, CompTIA, Windows, and Linux fundamentals. What once felt advanced turned out to be foundational.

A referral from someone he met during his CCNA course led to an opportunity at SoftLayer, later acquired by IBM, just as UK data centres were expanding.

“I joined on that wave and I helped build one of the data centres where I am right now.”

Certifications opened the door. Preparation secured it. Consistency kept it open.

There Is No Balance, Only Focus

When asked how he balanced study with full-time work, Tony’s answer was direct.

“It is not really about balance… it was just to focus and to prioritise what the goal was.”

Weekends were spent studying. Evenings were spent learning. It was not comfortable, but it was intentional.

Later in his career, growth became less about technical certifications and more about internal transformation. Between 2018 and 2022, he focused on improving business communication, reading leadership books, confronting limiting beliefs, and building confidence.

Public speaking, he admits, was one skill he wishes he had learned earlier.

“I’ve lost so many opportunities… I used to fear claiming my work and selling myself.”

The shift from technician to leader required more than technical depth. It required mindset.

The Industry Shift: Density, GPUs, and Revenue Per Rack

Having spent nearly a decade inside live environments, Tony has witnessed the evolution firsthand.

The biggest change has been density.

Traditional data centres required multiple rooms filled with commodity servers. Today’s AI and GPU-driven environments condense enormous compute power into high-density racks capable of generating significantly higher revenue per square metre.

“One rack… can give a lot of revenue rather than having a full rack of one-year or two-year servers.”

AI infrastructure has reduced physical footprint while dramatically increasing power density, cost per unit, and business impact.

And that changes the risk profile entirely.

“In the past you could break something and it was a $200 fix. Now you’re working on something that’s worth 1.5 million.”

The financial stakes are higher. The customer impact is greater. The margin for error is thinner.

The Hidden Constraint: Energy

While most AI conversations centre on models and capability, Tony believes the real constraint is infrastructure, specifically power.

“The amount of power they consume… it surpasses cities.”

As hyperscale and GPU-heavy data centres expand globally, energy availability is becoming a strategic bottleneck. This is one reason many AI facilities are being built in Scandinavia, where renewable energy sources such as geothermal power reduce both cost and cooling challenges.

Electricity demand is influencing national grid upgrades and pricing structures in certain regions. Governments are now thinking about data centre growth as part of energy policy.

AI is no longer just a software conversation.

It is an infrastructure conversation.
It is a power conversation.
It is an energy strategy conversation.

Engineers building these environments are operating at the intersection of compute, cooling, grid capacity, and national ambition.

AI Sovereignty and the Global Race

Tony offers a striking perspective on why governments are prioritising AI infrastructure.

“The countries now want AI and they want to be sovereign… it is a war.”

In his view, nations that fail to invest in AI and supporting the infrastructure risk falling behind economically and strategically.

This is not simply about model development.

It is about owning the capability that enables it.

Smaller Teams, Bigger Expectations

As hardware becomes denser and environments more automated, data centre teams are evolving.

Rather than large technician-heavy teams, organisations are building smaller, more cross-functional engineering units.

Tony favours generalists.

“People who can do anything, from cabling to customers to building the servers and talking to executives.”

GPU and HPC experience are in demand, but so are leadership, collaboration, and accountability. Smaller teams mean individuals must think more broadly and operate more independently.

The Most Powerful Skill: Mindset

When asked what matters most for long-term success in data centres and AI infrastructure, Tony does not name a platform.

He names mindset.

“The greatest skill is the mindset.”

Tools evolve. Technologies shift. Business models change.

But curiosity, ownership, adaptability, and resilience remain durable.

He also warns against becoming what he calls a button pusher.

“The tools don’t make you really viable. The judgement does.”

Understanding cause and effect. Asking why systems were designed a certain way. Anticipating downstream impact. These are the traits that separate operators from engineers.

Ownership, he says, is often underestimated, but it is what drives progression.

Looking Ahead: Space

When asked to describe the future of data centres in one word, Tony does not hesitate.

“Space.”

He believes energy and cooling constraints on Earth may push experimentation beyond the planet. With launch costs dropping significantly in recent years and early prototypes already tested, the idea of space-based data centres is no longer pure science fiction.

“AI becomes an energy system… as long as it’s inside Earth, there’s a problem with heating, cooling, and power.”

Whether orbital infrastructure becomes mainstream or not, the trajectory is clear. AI will demand new energy models, new cooling strategies, and new engineering disciplines.

One Piece of Advice for the Next Generation

If Tony had to give one piece of advice to aspiring engineers, it would be this.

“Put yourself where failure is expensive.”

In data centres, mistakes are memorable and growth accelerates under pressure.

“Failure is memorable and you learn really quick.”

We also, of course, finish off with that quick-fire round where his answers reinforced the broader message:

Public speaking is underrated

Certifications teach tools. Degrees teach positioning

Trust underpins all soft skills

Great engineers are comfortable with uncertainty

The Bigger Picture

AI may be reshaping computing.

But it is engineers, curious, adaptable, accountable engineers, who will determine how far it goes.

The future of infrastructure may involve denser racks, smaller teams, sovereign AI ambitions, or even data centres in orbit.

But the foundation remains the same. Mindset, ownership, and the willingness to step into environments where failure is expensive and growth is inevitable.

Tune In Now

Tune in now if you're an engineer, technologist, or infrastructure leader looking to understand where data centres are heading next.

From GPU density and energy constraints to sovereign AI ambitions and even space-based infrastructure, this episode explores the forces redefining the industry from the inside out.

🔗 Connect with Tony on LinkedIn here.