The best AI opportunities for investors may not look like AI companies
AI has become the investment theme of the moment. But for investors looking beyond the hype, some of the most interesting opportunities may have very little to do with building an AI company in the traditional sense.
Stanislav Shabaiev, founder of technology-focused investment holding company STNL Media Invest Holding, believes the bigger opportunity may be in businesses using technology to quietly transform established industries.
"AI is becoming a layer across the economy, rather than a sector on its own," says Shabaiev. "For investors, that changes the question. It is less about finding the next company with AI in its name and more about finding businesses that can use it to create a meaningful advantage."
That distinction is important. "There are very few industries left that are undigitised in any simple sense," Shabaiev adds. "What is left are industries with a digital shopfront and a manual back office. The customer sees the status of an order in an app, and behind that screen someone is moving data between three systems and picking up the phone to check it. That is where the value sits: not in giving an industry another interface, but in removing the manual work between the systems it already has."
The common thread is not a particular sector. It's the potential for technology to make a business more scalable, efficient and defensible.
"The model itself is available to everyone now," he says. "What creates defensibility is something else: data that only that company has, how deeply the product sits inside the customer's daily process and how expensive it would be to move away from it. If a product can be replaced over a weekend, there is no real advantage there, whatever technology sits inside it."
As capital pours into AI, the technology itself can sometimes become the story. "It is worth asking which budget the customer is paying from," he says. "A great deal is currently bought out of an experimental AI budget rather than the budget of the team that owns the problem. Those are different pots of money with different lifespans, and the experimental one is cut first. While the revenue comes from there, what we are looking at is curiosity rather than demand."
Yet investors still have to ask the same questions they always have: Who is the customer? Is there a genuine need? Can the business execute? Is there a path to sustainable growth?
"I mentally switch the product off for a week and ask what happens to the customer," he says. "If the answer is that it becomes inconvenient, it's a nice addition. If the answer is that work stops, that is a very different conversation."
These fundamentals matter more as AI moves from experimentation into everyday business operations.
"I'm less interested in pilots than in renewals," adds Shabaiev. "How many pilots became contracts, and how many contracts survived a second budget cycle. A pilot only tells you that a company is worth listening to. A renewal tells you that someone is willing to be accountable for it."
Shabaiev believes this next phase could favour companies that combine AI with deep knowledge of a particular industry. Rather than competing to build another general-purpose model, these businesses can apply existing technology to specific workflows, markets and customer problems.
"The winners may not be the companies that make the most noise about AI," he says. "They may be the companies that use it so effectively that AI almost disappears into the product."
This can create opportunities across Europe, particularly among founders building software and technology for industries that have traditionally been slower to digitise, such as manufacturing, construction and agriculture.
For investors, the lesson is straightforward. Look beyond the label.
The next generation of valuable tech companies may not describe themselves as AI businesses at all. They may simply be very good companies, using a powerful new technology to become much better at what they already do.
For STNL, that is exactly where the search gets interesting.
