AI Workshops

Hands-on experience with different AI tools in the research process – real      scenarios, not just demos

Ability to realistically assess the quality and reliability of AI results

Master targeted prompting – for optimal outcomes from AI tools

Develop your own ideas – how AI can support your daily work in a practical way

Building prompts instead of searching for them – the process with which you develop a fitting prompt for any task together with the LLM. The one competence that makes everything else obsolete.

Mastering meta-prompting – letting the LLM write the prompt you will then work with. The trick that makes it "click" for most people.

Prompt chains for complex analyses – breaking down large, vague tasks so that a comprehensible result stands at the end and not a plausible-sounding fantasy.

Knowing when what is enough – for which task a single prompt is sufficient, when a framework is needed, when the full chain. Saves you time on both ends.

Understanding why AI hallucinates – not as a bug, but as architecture. Whoever understands this, checks differently.

Source hygiene – your own data makes answers better. And wrong data makes wrong answers more convincing. How you deal with that.

A checklist for everyday life – before, during, and after prompting. Short enough that you actually use it.

Systematic skill and assistant development – the 4-stage maturity model for professional AI tools

Creating knowledge-based skills – how to integrate domain knowledge in a structured way instead of explaining it anew in every chat

Structured output formats – professional templates that travel within the skill and deliver consistent results

Robustness through testing – finding vulnerabilities and systematically fixing them before your colleagues do

In-prompt engineering – advanced techniques for reliable answers

Automation and integration – how to share skills within the team and embed assistants into real workflows

Strategic clarity instead of tool chaos – craft a clear AI vision for your team

Systematic potential analysis – identify where AI creates real value in UX work

Prioritized use cases – focus on applications with measurable impact

Implementation roadmap – move from experiments to established practice

Define success metrics – make AI impact tangible and trackable

Set the right framework – address data protection, infrastructure, and resources

The 2×2 map – two questions to ask about the task, and the automation boundary reveals itself

The hardness scale – recognizing whether your checkpoint really holds or is just a polite request

Verification that holds up – using data test sets, forced stops, and deterministic checks correctly

Reliability ≠ Validity – why a stable check is no proof of correctness

A judgment that moves with you – works equally well in Claude, ChatGPT, and Copilot

A process you truly understand – visualized, analyzed, and with its weak points marked

Methodological tools – BPA, BPMN-light, Muda: easily applicable, without certification overhead

A decision framework for technology – distinguishing between automation, AI, a combination of both, and "better leave it be"

A line of argumentation for stakeholders – a one-pager that prevents you from stumbling internally

Clarity on the sequence – optimize, standardize, automate. In exactly this order.