14 · AI and Automation
Prompt Engineering: Better Results with AI
The Prompt Engineering video course shows in 15 short videos, with no prior knowledge needed, how to word tasks for an artificial intelligence so that the results fit: from the six building blocks of a good prompt through examples and step-by-step thinking to guidelines and a helper that supports you in improving prompts. The videos are spoken in German, with German and English subtitles and a transcript and glossary for every lesson.
Who is it for?
For anyone who wants better results from an AI chat, even without technical knowledge.
The videos
Prompt Engineering
Introduction: Why good prompts matter
Here you find out what a prompt is, when good prompt engineering is worth the effort, and why the result is in your hands.
Naive prompt versus good prompt
Using a flyer for a painting company as an example, you see why a very short prompt often disappoints. With time, a clear description, an instruction and context, the result is usually better.
The six building blocks: role and clear instruction
You get to know the six building blocks of a good prompt and take a closer look at the first two: the role and the clear instruction, with examples from everyday working life.
Giving context and providing examples
When you explain to the AI what you need the result for and who will read it, the answer fits better. An example also shows it what the result should look like.
Input data, output format and a clean structure
Here is how you give the AI the material to work with, tell it what the result should look like, and structure your prompt clearly with headings and separators.
The prompt engineering process
Five steps to a good prompt: clarify the goal, write a first prompt, test it, improve it and repeat. With practice, your prompting gets better round after round.
Zero-shot, one-shot and few-shot: with no example, one example and several examples
Here is how you teach the AI what you expect with no example, one example or several examples, and when each version fits best.
Chain of Thought: letting the AI think step by step
Here is how to use the Chain of Thought technique: the AI shows you the way to the answer in small steps, and you keep the overview. The name became known through a research paper from 2022.
Tree of Thoughts: weighing up several solutions
Here is how you use the Tree of Thoughts technique: the AI shows you several ways to solve a problem, you compare what each one leads to, and you decide wisely. The name became well known through a research paper from 2023.
Prompt chaining: splitting a big task into small steps
Here is how you split a big task into small prompt steps: you stay in control and find mistakes more easily than with one single long prompt.
Bonus: small additions to your prompt that can improve answers
Three small additions for your prompt that can improve answers: step by step, the reason, and friendly encouragement. The effect varies, so the rule is: try it out.
Finding and using official guides
Here is how you find the official prompting guide for your AI chat and use its basic tips: clear instructions, a role and separators.
Using guides for different tools
Here is how you find the providers' guides, have the AI summarise English texts, and work in a focused way with your own texts.
The prompt improver: the AI asks questions and writes a better prompt with you
Here is how you let the AI ask follow-up questions about goal, tone and target audience, write a better prompt, and improve it further step by step.
More tips for clean prompts
Two simple tips for clean prompts: tell the AI what to do instead of only what not to do, and clearly separate the instruction from the material. Plus an encouraging end to the course.
Working material
Videos with German and English subtitles, a transcript, a glossary and small guides.
















