KI-Strategie-Loop® Example: A Machine-Building Company

Practice Story: How an SME Works Through the 7 Steps of the KI-Strategie-Loop (AI Strategy Loop)

Sample case: the company and people are fictitious; the steps show a typical course.

Company: Maschinenbau Muster Maier GmbH (40 employees)

An innovative family business with a strong customer focus, typical SME structures, and the wish to finally “do something with AI,” but with no clear idea of where to start.

Managing Director Anna Maier decides: “We need a systematic approach. No shots in the dark.”

Together with her leadership team, she works through the KI-Strategie-Loop®.

Perception – understanding what is happening around us

Anna invites her team to an “AI future breakfast.” They talk about:

• ChatGPT
• Predictive maintenance
• Automation
• new customer requirements
• the shortage of skilled workers

One employee says: “I had no idea how fast the industry is changing. We need to prepare for it.”

• Result: The team recognizes concrete trends and opportunities. For the first time, there is a shared picture of the situation.

Positioning – what do we actually want to use AI for?

The team formulates its first AI purpose:

“We use AI to improve service quality and to take the load off our employees.”

They define guidelines:

• AI must not replace anyone
• Transparency is mandatory
• Use data responsibly

• Result: A clear, understandable AI purpose, accepted by everyone.

Preparation – laying the foundations

The analysis shows:

• The data is scattered across Excel files
• Only two people know the machine history
• IT is solid but not AI-ready
• Employees are hesitant to engage

So the company decides to:

• start cleaning up data
• introduce a simple database
• send two employees to AI training
• set up internal “AI office hours”

• Result: A solid foundation, without overwhelming anyone.

Potential – discovering AI opportunities

Together, the team analyzes processes and customer problems.

They discover three relevant potentials:

1. AI-supported fault diagnosis in service
2. Automated quote generation
3. Prediction of machine failures (predictive maintenance)

Someone says: “If we pull this off, we’ll save a lot of time.”

• Result: Three concrete use cases, realistic and value-adding.

Prioritization – what is really worth it?

The team rates the use cases by:

• benefit
• effort
• data availability
• risk
• strategic importance

Predictive maintenance drops out for now (too little data). Automated quote generation comes in second. Winner: fault diagnosis in service.

• Result: Focus on ONE pilot project, not on ten construction sites.

Piloting – small project, big learning effect

The pilot project starts with a clear hypothesis:

“An AI can recognize 80% of the most common faults within seconds.”

They test:

• Real customer data
• Photos of machine problems
• Descriptions from service reports

Within three weeks, a working prototype is in place.

Service technician Tom says: “I never would have thought it could happen so fast. This really saves us time.”

• Result: Proof of concept, and enthusiasm in the team.

Performance – scaling success

After the successful pilot:

• The AI is integrated into the service process
• A short training session is introduced
• The process is documented automatically
• Error rate and processing time drop measurably after 3 months

The most important moment: a major customer says, “Your service is now faster than your competitors’.”

• Result: AI delivers real, measurable business value.

The loop closes, and starts over

Encouraged by the positive experience, the company wants to keep going:

• identify new potentials
• update the strategic positioning statements
• keep building data and skills

Maschinenbau Maier has turned AI into a continuous learning process, not a one-off project.

At the end, Anna Maier says:

“The KI-Strategie-Loop has made us a company that understands, uses and further develops AI. Not all at once, but step by step.”