Artificial intelligence is becoming increasingly important for small and medium-sized businesses. Those who ignore the technical possibilities it offers will soon be unable to keep up in the competitive race. And those who do not invest in time, but wait too long, risk falling so far behind that the gap can no longer be closed. On the other hand, however, the acquisition of AI by a company should be well thought out. Otherwise, there is a risk of spending a lot of money on solutions the company does not even need. «Many companies are currently buying a large-scale solution and then slapping an «AI» label on it because they believe their product will be worth more that way. But that’s the wrong approach,» says Jannik Reinhard, Head of AI at Epic Fusion.
Mr. Reinhard, what is the right approach when a company is considering implementing AI?
Do not start with the tool. Start with the business analysis. At Epic Fusion, we first want to understand how the work is done, which tools are already in use, and where AI actually adds value. AI must be integrated into the existing work environment. This way, we avoid creating new silos and place AI where it truly makes an impact. Our goal is to put AI at the heart of the company and make processes more efficient.
And once these questions are answered?
Then it comes down to three things: Governance, vision, and implementation. Governance comes first. Anyone who simply purchases AI without knowing what happens to their own data is taking an unnecessary risk and risks having their data transferred to American and Chinese providers. We must therefore clarify early on what data is being processed, where it is stored, and which providers we use.
And the other two points?
Second, you need a clear vision: What specifically is AI supposed to improve? Only then can you move on to implementation. Not every AI application is automatically worthwhile. It must deliver measurable added value for the company.
Is it advisable to proceed in stages, or is it better to implement everything at once? And what about software providers?
It is better to start gradually. We first identify a few areas and see how the AI solution works. At this stage, questions such as the following are important: Are employees using it? Does it save time? Where is there potential? We can then scale up these results and use them for the rest of the process. This is how we implement the solution step by step.
How important is it to involve employees?
Very important. AI is changing the way people work. That is why employees need to understand what is changing, how AI benefits them, and where human decision-making will remain. After all, AI can only be a success for the company if employees are on board. And because AI is constantly evolving, ongoing training is also part of the process.
How long does it take for the costs to pay for themselves?
Productivity and efficiency increase as soon as the first partial solution is implemented. And before long, the company can expand the solution more broadly, reduce the workload on employees, and operate much more efficiently. In addition, the company will need to rely less on external experts – e.g. for contract review – all of which saves costs. The key is not to view AI as a standalone tool, but as a lever for more efficient ways of working.