Thoughts on AI & Processes
Analyses, case studies and practical insights from AI consulting.
Why Good AI Results Need Source Work: Protection Against Hallucinations
Language models aren't search engines. They are masters of probability, not of truth. Anyone using AI professionally must understand the phenomenon of hallucinations and actively counter it – through clean source work and techniques like RAG (Retrieval Augmented Generation).
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.