Upgrading to an AI-powered SaaS model is not just a case of “adding AI” to an existing platform but adopting a completely new way of thinking. This activity alone is giving businesses across Europe a deeper strategic challenge than they imagined, because during each transformation process, they are sending hundreds of billions of dollars annually to vendors outside Europe, but in return control just 15% of their own cloud infrastructure.
Speaking at ChannelCon EMEA in London, Damir Tomicic, co-CEO and chief strategist at run.events, said the traditional way of digital transformation has to change in the AI era.
“AI is basically a four-year-old child, and sometimes it will do the right thing, sometimes the wrong thing. It knows everything we know, but just like a four-year-old about to start school, we are going to be sending it out into the world to learn and develop further. It needs to be guided properly, he said. “In the past 25 years every digital transformation I was involved in had the same pattern, but AI broke every rule I know.”
Tomicic said run.events initially tried adding AI to its existing SaaS platform, with the aim of creating a conference agenda for each specific person based on their interests, but it ended up taking too much time and left unhappy customers.
“Our customers were not happy with the end result,” he said. “We learned that if you just apply AI to existing systems without thinking—then it doesn’t work properly and you can end up with a lot of issues.”
Throwing more data at an existing system is counterproductive, he explained.
“A lot of garbage inside produces a lot of garbage outside,” he said. “We started to analyse what was happening in the system—not just with the software but with the services as well and concluded that AI brings you the answer you are looking for but requires different data. How you rebuild the software has to change—you need to move into a different space and it’s a new way of thinking.”
To see success, data needs to be separated into three groups, he explained: Experience, automation and intelligence. In short, separate compute from cognition.
“We used tools offline to prepare the data and doing it offline means you can purify your data to help you make better decisions based on quality data.”
For MSPs, he explained, this could mean thinking about how they apply their services and creating a stronger corporate identity. For example, who are they? Are they the perfect operator with the perfect tools and the perfect services?
Striving for perfection is also to be avoided, he warned.
“As human beings we are not perfect, and you have to force AI to make mistakes—this is called reasoning. Otherwise, you end up with a result that is just robot speak.”
Thinking about where data is stored is also something firms must consider for the future, Tomicic warned.
“In Europe we own just 15% of our data infrastructure. This is dangerous because data is moving to a political place and we need to start thinking more about where it is stored. How sure are you that you can access your stored data?”
He added that when undergoing digital transformation with AI you must question everything all the time.
“AI is still that kid that has to be questioned—how much of its development has been led by you and how much has AI developed itself? This is something you have to bear in mind,” he said. “At the end of the day it is the way we think about AI that has to change, and one of the hardest things to do is delete all past assumptions and start again with a different mindset.”
GTIA Members: Check out AI readiness resources on the Member Portal.
