The arrival of AI agents is causing a rift: do you unleash autonomous AI agents on your business processes, or is it better not to? The Hasselt-based software company Lowi is going directly against the agent hype.
AI agents distinguish themselves from chatbots because they can perform tasks autonomously. Although these agents haven’t been around in the AI landscape for long, they are being embraced by the majority of companies. Lowi, on the other hand, opts for a Flemish approach.
“There is currently a kind of gold rush going on around fully autonomous AI agents. More and more companies are starting to blindly trust AI to automate their organization. I simply cannot wrap my head around that,” begins Klaus Dillen, CEO at Lowi.
Against the hype
The Hasselt-based software company Lowi is charting a different course against the current of the agent hype. Dillen sees how companies are increasingly relying blindly on autonomous AI to automate their processes. He finds that risky.
AI models do not deliver the exact same outcome twice.
Klaus Dillen, CEO at Lowi
“AI models work on the basis of probability and are unpredictable. They do not deliver the exact same outcome twice,” he says. According to him, this is guaranteed to go wrong.
Deterministic approach
Instead, he is a proponent of a deterministic approach. “With Lowi, we try to make processes predictable and reliable in a deterministic way. Processes are mapped out step by step in a way that people understand, so that employees always know what is happening and can intervene if necessary.”
Top-down intelligence, bottom-up reliability
Lowi looks at companies in their entirety for this. “Our platform maps out the strategic position, the competitive landscape, the internal processes, and the underlying data sources using AI,” Dillen explains. That total picture, also called top-down intelligence by Dillen, forms the basis for targeted improvement proposals.
The automation does not happen all at once, but step by step: Lowi asks targeted questions to understand and optimize a process, without promising that everything is fully automated from day one.
Lowi asks targeted questions to understand and optimize a process, without promising that everything is fully automated from day one.
Klaus Dillen, CEO at Lowi
“Once a process has been mapped out, it is converted into a predictable and insightful whole; we call that bottom-up reliability.” Instead of unpredictable AI that decides independently, the platform chooses a deterministic approach where processes run the same way time after time. People remain involved: they can follow, monitor, and adjust the processes where necessary.
AI the Flemish way
According to Dillen, there is a Flemish philosophy behind Lowi. According to him, AI is not a goal in itself, but a means to achieve results. “We are driven by technology, but with a smart perspective and always with an eye for quality. When the hype has passed, Lowi will remain,” he states. “Our down-to-earth, results-oriented mentality is the polar opposite of the ‘magical’ thinking surrounding AI.”
No magic
“Automation should not be magic, or something that people do not understand or cannot control,” Dillen states. People must understand how a process works, have control over it, and be able to intervene where necessary.
Lowi recently raised 2.7 million during a funding round. The fresh capital will mainly be used to accelerate the commercial rollout in the Benelux and subsequently Europe. “We continue to build on the further optimization of Lowi, and we are also focusing on international growth,” Dillen concludes.
Are you using AI agents within your company? ITdaily is currently gauging the use of AI agents at organizations. The results will be announced soon.
