1. What is "Prompting"?

A prompt is simply the message you send to the AI in the chat.

The clearer your request, the better the result.
Think of it like briefing a colleague: What should they do, with what information, in what style?

2. The Basic Rules for Good Prompts

Be specific: "Write a customer email" is too vague. → "Write a polite email to a customer whose invoice is 14 days overdue."

Include context (such as the email you're responding to).
You can also upload files for the AI to use (paperclip symbol).
What's the situation, who's the target audience, what's the goal?
Add additional instructions as needed, such as:
- A role: "You are a salesperson" or "You are customer support"
- "Reply in bullet points"
- "Be friendly but firm"
- "Reply in English"

3. First Tasks the AI Can Take Off Your Hands

The KI-Workplace can save you time and effort on many everyday tasks and provide input.

This is especially worthwhile when working with sensitive texts and data.

The KI-Workplace is completely data-secure. You can enter information there that is personal, customer-related, patented, or confidential for other reasons. No one can read your chats, not even other people on your team, your admin, or us at Adacor.

It's like getting to know each other: Over time, you'll get better at finding out how to "press the right buttons" with the AI to get the best results.

But even on your very first use, you can achieve success with the right prompts. Here are some examples:

📧 Writing Emails

Customer Email: "Write a friendly response to the following customer email: [insert question]. Reply briefly, professionally, and with reference to our website [URL] for further information."

Sales Email: "Suggest an outbound email to target group X for the following product: [insert product information]."

Rejection: "Write a friendly rejection email for the following offer: [insert offer]. Express appreciation and explain the rejection due to current budget constraints."

Circular: "Write a friendly email to staff explaining the key points of the new policy and encouraging them to read it: [insert policy]."

Reminder: "Draft a reminder email to a customer who is four weeks behind on the following invoice: [insert invoice]. Be friendly but binding and professional in tone."

📝 Analyzing or Generating Texts

Summarizing Presentation: "Summarize the most important points and recommendations of the presentation in a few bullet points so my manager can quickly understand them: [upload presentation]."

Meeting Notes: "Summarize this MS Teams meeting transcript by briefly listing the topic progression in bullet points and listing all decisions and to-dos we agreed upon: [insert transcript]."

Language Editing: "Check the following text for clarity, errors, and conciseness. Give an assessment and provide me with a revised version of the text: [insert text]."

Writing Project Report: "Based on the following project plan, write a report to stakeholders with the most important key data: [insert project plan]."

Translation: "Translate the following text into [insert language], using formulations that fit the business context: [insert text]." (This works best between German, English, and widely spoken languages.)

Compare Documents: "Compare the following two documents and list differences in content, format, language, and quality: [insert/upload two texts]. Give a recommendation on which text is better suited for our website."

Modify Text Sections: "Remove all personal data such as names, email addresses, and addresses from the following text and replace them with placeholders: [insert text]."

Coding 🤖

Python Script for Logfile: "Here's an excerpt from our internal logfile: [insert log file or text excerpt]. Write me a small Python script that counts how often each error type occurs."

JavaScript Snippet for Table: "Here's a short list of internal employee data: [insert text or file]. Write me a JavaScript snippet that displays this data as a table in an HTML page."

✅ To-Dos & Checklists

Structure Conversation: "Analyze the following email thread and create a step-by-step checklist for an upcoming meeting about it: [insert email thread]."

Derive Tasks: "Create a list of all to-dos for our live event: [insert event concept]."

📊 Ideas & Brainstorming

Quiz for Learning Material: "Take the learning material and create a quiz of 5 questions that increase in difficulty. Give four possible answers for each question, only one of which is correct, and mark the correct answers. [insert learning material.]"

Event Strategy: "Attached is the agenda of our last hackathon, as well as collected feedback. Give me 5 creative ideas on how we can make the next hackathon more exciting. [insert agenda and feedback.]"

🔍 Start Research

Explain New Topic: "Explain to me in simple terms the most important points of the new [insert topic, e.g., law or technology] and why it's relevant for medium-sized businesses." (Activate the web search button for this so the AI researches on the internet.)

Research Software: "Recommend 2-3 online survey tools for the following questionnaire, list their pros and cons, and state their prices: [insert questionnaire]." (Activate the web search button for this so the AI researches on the internet.)

4. Continue the Conversation

You can refine and further develop the AI's responses by asking additional tasks or questions.

Examples:

"Write a shorter version of that."
"Make the tone of the text friendlier and more informal, and use 'you' informally so it better matches my usual style."
"Create a shopping list with things we need for such an event."
"Explain your last answer in more detail."
"You made an error at point X, please correct that."
"Convert the answer to Markdown."
"Rewrite the text for Gen Z and add emojis."

👉 Getting Started Tip: Try the examples above and then vary them a bit. You'll quickly notice how slightly modified prompts can also achieve increasingly suitable results.

Do you have questions or feedback about this? Feel free to write to us at [email protected].

Have fun and success!

Your KI Workplace Team