💬 Day 2: Prompt Like a Pro – Stirring the Perfect Query

Extra Learning Materials for the keen beans.

Hello again!

Today we’re diving deeper into one of the most important ingredients in our AI coffee analogy: Prompting!

In our coffee analogy, we’ve looked at how training data is the beans and fine-tuning is the brewing process. But it’s prompting that acts as the recipe—it’s what decides the final taste and experience.

Without a recipe, even the best beans and brewing method can’t deliver a great cup. In the same way, a well-crafted prompt is what transforms tools like ChatGPT from generic systems into useful, relevant, and personal solutions for your tasks.

🎯 Why Prompting Matters 🎯

Prompting is the most underestimated part of working with Generative AI, and the single best starting point for anyone looking to understand it.

Whether you’re using AI as a chatbot, assistant, automation, or agent—prompting is the foundation. It’s how you instruct the AI, shape its tone, and guide its output.

The better your prompt, the better your results.

And like any good recipe, it can be refined over time.

The prompts you provide directly shape the quality of your AI’s responses. Whether you’re training a chatbot, collaborating with your AI assistant on strategic reports, or automating routine tasks—the clarity and precision of your instructions truly matter.

Let’s look at 2 methodologies we can use for effective prompting:

🍯 Precision Prompting Method 🍯

• ✅ Clarity & Context:

  Tell the AI who it is, what you need, and why it matters.

  Example: Act like a student advisor at University of AI in Wonderland. Help me draft a clear and helpful information email on visa deadlines, accommodation options, and programme specifics for our incoming international students who will need guidance in their first weeks here.

• 📊 Incorporate Real-World Data:

  Use real institutional FAQs, policies, or past messages to make outputs more accurate and trustworthy.

  Example: Make sure to tailor your response based on our attached institutional policies.

• 🗣️ Use Specific Examples:

  Share templates or sample conversations to help AI maintain your institution’s unique tone and style.

  Example: Ensure to maintain our tone of voice as per the attached past communication. [attach files]

• 🔄 Refinement:

  Regularly adjust your prompts based on user feedback to continuously improve accuracy and relevance.

  Example: Adjust your writing style to make the message friendlier, whilst keeping a professional tone.

🫘 Strategic Prompting Model 🫘

We can also think about prompting in a slightly different structure, keeping in mind four main ingredients you should also consider when writing a prompt:

• Goal: What response do you want from the AI chatbot?
• Context: Who do you need this for and what is their background?
• Source: What information sources should be used?
• Expectations: How should the AI respond to meet your expectations?

✨ Today’s Action Step ✨

Practice: Choose one of the methodologies above. Create and refine a prompt for a routine task you regularly perform. Notice how improvements in your prompt’s clarity positively affect your AI’s performance.

Tomorrow, we’ll explore how your well-crafted prompts become the backbone of one of AI’s most practical tools: the AI assistant.

Enjoy your coffee break,

Cato ☕✨