Thoughts on AI & Processes
Analyses, case studies and practical insights from AI consulting.
AI Literacy Under the EU AI Act: From Law to Team Capability
The EU AI Act contains a far-reaching provision in Article 4: organizations must ensure their staff has sufficient AI literacy. What sounds at first like administrative burden is actually the most important prerequisite for safe and productive AI use.
Local AI Models and Isolated Setups: Maximum Security for Sensitive Data
When patient data, business secrets or highly sensitive strategy documents are involved, cloud-based AI solutions hit their limits – not technically, but regulatorily and ethically. The solution exists: local language models today offer powerful AI without a single bit leaving your own network.
Microsoft Copilot in Daily Work: Realistic Assessment Instead of Blind Trust
Microsoft Copilot is often marketed as the ultimate tool for the modern workplace. Anyone using it in practice quickly notices: it isn't an autopilot, it's an assistant that needs guidance. When Copilot shines, why your data structure decides success – and where the system's limits lie.
AI Policy for Small Organizations: Simple Rules Beat Thick Manuals
Many organizations hesitate to use AI because they think they need an extensive legal framework first. Especially for small businesses, associations and nonprofits, a pragmatic, concise policy is often safer than no policy at all. It gives the team something to lean on – and creates clarity without paralyzing.
What Leaders Really Need to Know About AI
Leaders don't need to become prompt experts. But they must be able to assess benefits, risks, limits and responsibilities. A compact orientation framework for decision-makers who want to anchor AI responsibly in their organization.
Small Organizations Don't Need a Big AI Strategy
Associations, social-services organizations, small businesses and initiatives often hesitate because they think they first need a comprehensive AI strategy. That's a misconception. What small organizations really need: a few clear use cases, simple rules and a practical training.
Knowledge Management as the Underrated AI Foundation
AI can only use knowledge well if it's findable, current and understandable. Many organizations skip this step and are then surprised by disappointing results. Why folder logic, documentation routines and clear responsibilities are the most important prerequisite for meaningful AI use.
From Idea to AI Work Template
Every team has recurring tasks suitable for AI support. But how does a vague idea become a usable work template? This article shows the path from task identification to finished template – with concrete examples.
How AI Training Drives Real Change
Good AI training focuses on transfer into everyday work. What makes the difference: task focus, embedded data protection and sustained follow-up. Not the most spectacular demo – but a methodical path to real competence.
Understand the Process First, Then Apply AI
AI accelerates clear workflows and exacerbates unclear ones. Anyone automating a chaotic process simply gets chaotic results faster. Why a process analysis before introducing AI decides whether the investment delivers or evaporates.
Prompt Engineering Is Not a Magic Trick
Good prompts aren't magic. They are structured work instructions with goal, context, role and quality criteria. Why professional prompt systems matter more than creative one-off prompts – and how they actually work in organizations.
The EU AI Act as a To-Do List for Organizations
The EU AI Act isn't a reason to avoid AI. It's a reason to introduce AI deliberately. What this regulation means for organizations in practice – and why AI literacy, risk awareness and clear usage rules matter more than legal perfection.
Hello World
For years, I have been closely following the latest developments in global digitalization and AI trends. How these directly impact our environment, which risks we must be aware of, but above all how we can truly benefit from them—this is what I will regularly write about here.
Using AI Responsibly with Sensitive Data
"We can't use AI because our data is sensitive." This sentence blocks many organizations. But sensitive data does not automatically mean an AI ban – it depends on which information goes into which system and how the conditions are regulated.
Why AI Adoption Rarely Fails Because of Technology
Most AI projects don't fail because of missing technology. They fail because of unclear goals, unsuitable processes and lack of acceptance within the team. Why organizations should clarify tasks, data and responsibilities first – before selecting a tool.