Prompting Tips

In this guide, you will receive valuable tips on how to formulate effective prompts and thus get the most helpful answers from Alan.

1. Clarity and Specificity

Formulate clear and specific questions to avoid ambiguities and obtain accurate information.

Positive Examples:

• "What are the main features that distinguish Python as a programming language?"
• "How can artificial intelligence increase efficiency in the logistics industry?"

Negative Example:

• "Tell me something about Python." (too vague)

2. Detailed Context

Provide enough background information so that Alan can understand the scenario and deliver a relevant answer. If you have documents available as background information, you can also provide them using the Chat with Files feature. This way, Alan can focus on the content even more precisely.

Positive Examples:

• "As a project manager starting a new software project, what steps should I take to ensure all team members are informed?"
• "I plan to open a small café in a busy downtown area. What marketing strategies would be most effective to attract local customers?"

Negative Example:

• "How do you start a project?" (without specific context)

3. Desired Format

Specify the format in which you want the answer (e.g., list, report, bullet points, summary, JSON).

Positive Examples:

• "List five advantages of cloud computing."
• "Create a JSON list of all Formula 1 world champions with their name, country, number of championships, and the year of their first title."

4. Output Length

Define the desired length of the answer (e.g., "3 paragraphs" or "200 words") to avoid responses that are too short or too lengthy.

Positive Examples:

• "Summarize the French Revolution in approximately 200 words."
• "Explain the advantages of electric cars in three paragraphs."

5. Level of Detail

Indicate how in-depth the answer should be, from a simple overview to a detailed analysis.

Positive Examples:

• "Explain the basic concepts of quantum physics so that a five-year-old can understand them."
• "Describe in detail how an internal combustion engine works."

Negative Example:

• "Explain quantum physics." (without specifying the desired level of detail)

6. Tone and Style

Choose the desired tone and style for the answer (e.g., formal, informal, persuasive, factual).

Positive Examples:

• "Explain the benefits of recycling in an informal tone."
• "Write a persuasive text about the importance of data privacy."

7. Examples and Comparisons

Ask for examples or comparisons to make complex concepts more understandable.

Positive Examples:

• "Explain the difference between symmetric and asymmetric encryption using a clear example."
• "Use a simple analogy to explain how neural networks work."
• "Compare blockchain technology to a traditional ledger to explain how it functions."

Negative Example:

• "Explain how neural networks work." (without requesting a comparison)

8. Step-by-Step Explanations

Instruct Alan to explain his thought process step by step. This helps to understand the logic behind the answer and to solve complex problems systematically.

Positive Example:

• "I need to prepare a workshop on using AI in our company. Help me plan it step by step."

Negative Example:

• "Prepare a workshop on using AI in our company." (without requesting step-by-step planning)

9. Iterative Refinement

Start with a general question and refine your query based on the responses received to obtain more in-depth information.

Positive Example:

• "Explain the basics of machine learning."
  • After the answer: "Can you go into more detail about supervised learning?"

10. Role-Playing and Perspective Shifts

Ask questions from the perspective of different roles or individuals to gain different viewpoints.

Positive Examples:

• "How would an experienced IT manager explain the advantages of cybersecurity measures?"
• "Argue from Steve Jobs' perspective about the importance of innovation in modern companies."

11. Use of Constraints and Specifications

Provide specific constraints or requirements to tailor the answer better to your needs.

Positive Example:

• "Explain the theory of relativity without using mathematical formulas."

12. Using Examples in Your Request

Incorporate examples or patterns in your question to give Alan a better understanding of your expectations.

Positive Example:

• "Write an email to a customer who is dissatisfied with our service and offer a solution. For example: 'Dear Mr. Smith, we are sorry to hear that...'."

13. Allow Feedback and Follow-up Questions

Encourage Alan to ask follow-up questions or seek clarifications if something is unclear.

Positive Example:

• "If anything in my request is unclear, please let me know before you answer."

14. Enriching Requests with Markdown

Insert text with Markdown formatting into the chat input field.

Since Alan understands Markdown, relevant information about the structure of the text (e.g., headings or highlighted keywords) can be taken into account when generating answers. This allows Alan to provide more accurate and helpful responses that reflect the structure of your request.

You can conveniently convert text from your clipboard to Markdown in Alan and then insert it into the chat input field. Click on the plus (+) button in the input bar and click "Paste formatted text". Alternatively, you can select the "Paste text as Markdown" action in the action bar or use the keyboard shortcut v.

Positive Example:

Summarize the following email.
List the most important points and keep the summary brief and concise.

# Email from Mrs. Schmidt

## Subject:

Protocol of the project meeting for the new website

## Content:

Dear all,

here are the most important **decisions** from our project meeting:

- **Website Structure**:
  - **Start Page**: Overview of our company
  - **Product Page**: Detailed information about our products
- **Design**:
  - **Color Scheme**: Blue and White
  - **Font**: Open Sans

### Next Steps:

1.  **Website Development**: Start of website development
2.  **Content Creation**: Creation of content for the website

Best regards
Mrs. Schmidt

15. Referencing Uploaded Files Precisely

When you chat with files, Alan is aware of both the filenames and the order in which you uploaded them. You can use this to make targeted comparisons or identify differences between documents.

Positive Examples:

• "Compare the file Contract_2024.pdf with Contract_2023.pdf and list the differences."
• "What information from the second uploaded file is missing in the first?"

16. Referring to Document Structure While Chatting with Files

When chatting with files in Alan, the system understands the internal structure of the documents - such as sections or chapters. You can tailor your questions to focus on specific parts of an uploaded document.

Positive Example:

• "Summarize the third section of the uploaded document."
• "What recommendations are made in the final chapter of the file?"

Negative Example:

• "What’s on page 5 of the document?" (Note: Chatting with Files doesn’t interpret page numbers - only internal structure.)