Implementing AI solutions? You probably started “From the wrong end”
The corporate world has fallen under the artificial intelligence (AI) spell: terms such as “optimization,” “automation,” and “efficiency” dominate boardroom agendas, while IT teams face growing pressure to rapidly “implement AI.” But many forget that AI is not a rocket – it’s a train, and it doesn’t move… without tracks.
What are those tracks? Why do even the most advanced AI initiatives often fail without them? And where should you start laying them to make sure your data is ready for the journey? Insights on this topic are shared by Martynas Spokas, Altic IT Partner and Data Platform Architect.
Tracks for the train or when it’s time to move beyond Excel
Many businesses see AI as the ultimate goal – a smart solution that will solve problems, increase profits, or speed up processes. But they forget that AI cannot function on its own – it needs a solid foundation. That foundation is data platforms: the infrastructure without which no AI project has anywhere to “run.”
A data platform is a technological base that collects, processes, and unifies data from all of a company’s different systems: accounting, sales, customer service, procurement, inventory, logistics, or warehouse management. It not only standardizes the data but also prepares it for further use – analysis, forecasting, automation, and ultimately AI models.
Many companies try to replace this foundation with Excel – until it crashes under the amount of data. Others attempt to connect “prettier” analytics tools directly to business systems, which were never designed for that direct purpose. Then the problems begin: systems that malfunction, mismatched numbers – “different versions of the truth” – and growing internal mistrust of the data. This is a clear sign: the train has started moving, but the tracks underneath it are not yet laid.
Where do you start laying those tracks?
Still, a train won’t move just because you lay down tracks – you also need to know where they should lead. In other words: where do you begin your data journey?
Most business leaders today understand that data is no longer just a technical add-on, but the foundation of strategic management. They set ambitious goals for IT departments, expecting more analysis, insights, and speed – yet often see little real benefit from their data.
Because of disconnected systems and lack of unified access, data becomes fragmented and unreliable. Decisions are then made based on intuition rather than facts. Or data is used only for reporting – like driving a car while staring solely in the rear-view mirror: you see what already happened, but can’t react in time to what’s ahead.
The first step to unlocking real value from data is organization-wide access to the same, unified, processed information. We’ve often seen sales and finance managers answer the same question with completely different numbers – and both were “right.” Why? Because they relied on different sources.
A unified data platform eliminates that confusion and aligns the whole organization: it helps identify sales opportunities more accurately, calculate real customer profitability, and see where processes can still be optimized.
Once the foundation is solid, you can move further – apply forecasting, automate decisions, and of course, implement AI. Later, you can even turn advanced data solutions into entirely new products – or monetize the data itself.
Or maybe you don’t even need the train – but the tracks are still useful
We talk about data platforms in the context of AI because they prepare the data for AI solutions. But the truth is… not every company needs AI today. For some, it’s too early; for others, it’s simply not their most pressing issue. Still – it’s worth laying the tracks.
Why? Because even if your “train” isn’t on the horizon yet, the tracks can serve other purposes. For example, ensuring every department works from the same data, that customer profitability is calculated consistently, that sales see the same picture as finance. In other words: ensuring the company stops “working on gut feeling” and starts making decisions based on real, trusted information.
We’ve seen many times how a well-implemented self-service reporting system can bring more clarity and value than trendy AI projects launched without clear goals, maturity, or real business benefit. Many companies still rush into AI “for the hype” or fear of falling behind, proudly saying they “use AI daily” – but without clear objectives and a solid foundation, those initiatives deliver little to no results.
A data platform is not just the base for AI. It is the nervous system of the entire organization – functioning even when AI is still a distant station. And sometimes, once you start working with clean, centralized, and reliable data, you realize you can travel much farther – even without the train.
