WEBSI returned for its 15th edition last week, and Infinum was part of it – with a talk on how to take AI transformation out of the board rooms and into the organizational workflows.
Websi has been Slovenia’s leading digital competition for 15 years, bringing together marketing and IT companies from across the country to showcase the best of what the local industry can build.
Projects compete across multiple categories – digital products, campaigns, platforms, AI, and more – and the audience is exactly the kind of community that makes an event like this worth showing up for: practitioners who take the work seriously.
This year’s entries made that clear. The ideas were sharp, the execution was serious, and the standard across categories was genuinely high. To the organizers, the competitors, and everyone who put their work up for scrutiny – congratulations.
Our contribution, however, came from Željko Plesac, Infinum’s Partnerships Director.
Željko joined Infinum 13 years ago as an Android developer, spent a decade leading mobile banking and neo-banking projects in the financial sector, and now manages the partnerships and client relationships at the centre of our business – work now centered on AI transformation.
Naturally, he started in-house, bringing AI into Infinum’s non-production teams, where he figured out what actually makes adoption stick.
His topic at WEBSI: what AI transformation actually requires – drawn from the programme Infinum ran on itself before taking it into regulated industries.
AI transformation used to mean discovering the path
Most organizations find themselves in the same position. The board has asked for an AI strategy. Twice. Engineers are using tools that legal hasn’t fully approved. A few enthusiasts are pushing adoption while everyone else waits.
The ROI is theoretical even after the licences are paid. No one knows any metrics on how the AI transformation is going – only that it’s “going well”.
Most organizations find themselves in the same uncomfortable position: the ROI is still theoretical, and no one knows any metrics on the AI transformation – only that it’s “going well”.
Infinum knows this because we lived it.
Starting in 2023, we went from unstructured experimentation – with one firm rule: don’t compromise client data – through internal process testing in 2024, to a company-wide rollout in 2025 built on defined engineering ways of working and AI champions in every team.
Today, over 90% of Infinum’s engineers have been through structured AI training sprints. Getting there took over two years, a team restructuring, and a rewrite of how we work. None of the hard parts was the technology.
In 2026, the rollout took the shape of custom training sprints, unified tooling, central governance, and forward deployed engineers embedded across the business to drive adoption from the inside.
Getting there took two years, a team restructuring, and a rewrite of how we work. That doesn’t mean the work is finished.
Adoption is still uneven in places – but now we know exactly where it stalls, and why. That’s the actual outcome of building the right infrastructure: not a solved problem, but full visibility into the one that remains.
“When we started, best practices lived in a couple of people’s heads. Every engineer set their own rules, and leadership couldn’t answer whether any of it was working,” – explained Plesac.
“Now we have one company position on tooling and processes, and a dashboard that shows us exactly where adoption still has gaps.”
Now there’s a route
What lasting AI transformation really needs isn’t the tools – it’s governance, ownership, and the ability to measure.
That’s the foundation Infinum built internally, and the same one we now help clients build in industries where change has to be careful by design.
The organisations that get AI transformation right aren’t the ones that move fastest. They’re the ones that put governance and ownership in place before they scale adoption – and that applies as much in regulated industries as anywhere else.
The programme Infinum ran on its own is now the one we take to clients. It runs in three phases.
Clarity: Before we recommend anything, we listen and look – speaking with leadership, reviewing how AI is actually being used in practice, and giving an honest picture of where the organisation stands.
Change: This is where the work happens. We embed inside teams and work on real projects, not presentations. By the end, people are doing things differently.
Capability: The real measure of a successful transformation is what remains after we leave. We hand over everything the organisation needs to keep improving on its own.
AI transformation is finding its purpose across every industry — not just the ones that ship code fastest. That’s the case Željko made in Ljubljana, and we hope it landed.
WEBSI – see you again.