There’s no shortage of ambition around AI. In fact, GDS data shows that 99% of tech leaders are now actively working on some form of AI initiative. The intent is clear. The investment is growing.
But outcomes are inconsistent.
At the Tech Leaders Europe Summit, Giacomo Fiocchi offered a perspective that cuts through the noise: effective AI isn’t accidental. It’s architectural, and it needs an AI data strategy.
The disconnect between ambition and execution
Most organisations don’t struggle with why AI matters. They struggle with how to make it work.
There’s a tendency to focus on outputs—use cases, automation, ROI. But as Fiocchi points out, that’s the visible layer. The reality is more foundational:
“Everyone looks at the return on investment… but there’s a boring part behind. You need to understand your strategy starting from the lowest layer—the foundation.”
This is where many AI strategies stall. Not because the vision is wrong, but because the groundwork hasn’t been done.
AI is only as strong as the data beneath it
The comparison is simple, but effective: building AI without a data strategy is like building on unstable ground.
Before any model is deployed or scaled, leaders need clarity on three things:
- Where data lives
- How it is structured and governed
- What data actually matters to the business
These aren’t technical afterthoughts. They are strategic decisions that shape everything that follows.
And yet, this is often the least visible—and least prioritised—part of the journey.
Infrastructure is a strategic choice, not an IT task
Cloud strategy, data architecture, and storage decisions are often delegated deep into the organisation. But their impact is enterprise-wide.
Fiocchi highlights that these choices need to be made early—and deliberately:
- What is your cloud approach?
- How interoperable are your systems?
- Can your infrastructure support scale, not just experimentation?
Without alignment here, even the most promising AI initiatives become isolated pilots.
From experimentation to execution
What separates organisations that scale AI from those that stall isn’t access to tools or talent. It’s discipline in design.
That means:
- Treating ai data strategy as a board-level priority
- Aligning infrastructure decisions with long-term AI goals
- Accepting that the “boring” work is often the most critical
Because ultimately, AI success isn’t defined by what you build—it’s defined by what you can sustain.
A more grounded approach to AI
The conversation around AI is maturing. The focus is shifting from possibility to practicality.
And that’s where real progress happens.
Not in chasing the next use case, but in strengthening the layers that support them.
Because the quality of your AI outcomes will always reflect the quality of the foundation beneath them.