KI Fahrstunde: Prompting & LLMs

Learn to craft reliable LLM prompts beyond trial-and-error. We cover fundamentals (Goal, Context, Format), patterns (Personas, Few-shot, Checklists), and common pitfalls. Hands-on practice with your own use cases (emails, analyses, drafts) leaves you with a reusable prompt toolkit for daily work.

Sign up
Large studio

LLMs have arrived in our daily lives, yet the results often feel like a roll of the dice: sometimes helpful, sometimes way off the mark. In this AI Driving Lesson, you will learn how to craft prompts that deliver reliable outputs tailored to your specific use case.

We start with a compact crash course on prompting fundamentals (Goal, Context, Format), best practices (Personas, Examples, Constraints), and typical stumbling blocks (Hallucinations, vague requirements, missing verification steps).

Following the theory, we dive straight into a guided exercise: you will apply prompting patterns, test variations, and learn how to evaluate and refine results. In the main segment, you will work on your own real-world cases—such as emails, analyses, conceptual drafts, or Q&A for long texts—supported by the trainer as your "co-pilot" and through peer learning within the group.

By the end, you will leave with concrete outputs and a small "toolkit" of prompt patterns that you can use immediately in your daily routine.

Target Audience

Anyone with initial experience using AI/LLMs who wants to move beyond "trial and error" toward confident, goal-oriented application.

Prerequisites

  • No technical prior knowledge required.
  • Laptop (Wi-Fi enabled).
  • Recommended: Your own LLM access (e.g., ChatGPT, Claude, Copilot) and 1–2 specific use cases or examples from your daily work.

What you'll learn

  • Structure prompts clearly (Goal, Context, Format, Constraints) and avoid common mistakes.
  • Apply prompt patterns (e.g., Personas, Few-shot prompting, Checklists, Iteration) effectively.
  • Critically evaluate LLM results for quality, completeness, and plausibility, and improve them.
  • Translate your own use cases into effective prompt strategies and optimize them iteratively.
  • Create reusable prompts/“recipes” that you can implement directly in your daily workflow.