Lesson 2.2 – Prompt Engineering

Unit 2 · Lesson 2.2 – Prompt Engineering | Knowledge Pillars · AWS AI Explorers

Lesson Objectives

By the end of this lesson you will be able to:

Apply prompting techniques: context-setting, constraints, role assignment, and examples
Iterate on prompts systematically to improve output quality
Build a prompt template that produces consistent results for a specific task
Evaluate outputs for accuracy, bias, and relevance

Four Prompting Moves

A strong prompt is not magic; it is built from a few clear moves. Use them together and your results get far more consistent.

🎯

Context-setting

Say what the task is for and who it is for.

📏

Constraints

Set length, style, and format.

🎭

Role assignment

Tell the AI who to be.

🧩

Examples

Show a model of what you want.

Put them together: You are a study coach. Explain photosynthesis to a 9th grader in 3 short steps. That single line uses a role, a constraint, and context.

Iterate to Improve

Your first prompt is a draft. The skill is iterating: read the output, change one thing, and try again. Better prompts produce better, more consistent outputs, and the link is direct.

Change one thing at a time. If you change everything at once, you will not know what helped.

Evaluate Your Output

Never use an output just because it looks polished. Check it against three questions before you trust it.

Check	Ask yourself
Accuracy	Is it actually true? Did I verify the key facts?
Bias	Is it fair? Does it leave anyone out?
Relevance	Does it answer what I actually asked?

🧑‍💻 Activity: Build a Prompt Template

Activity

Make a reusable prompt

A template is a prompt you can reuse for a task you do often. Build one and test that it gives consistent results.

Pick a repeated task. For example: writing a friendly reminder, or summarizing an article.
Write the template. Include a role, the task, the context, and the format.
Test it twice. Use it on two different inputs and check the results are consistent.
Evaluate. Check both outputs for accuracy, bias, and relevance, then refine the template.

Lesson Summary

Idea	Why it matters
Four moves	Context, constraints, role assignment, and examples.
Iterate	Change one thing at a time to improve the output.
Quality link	Better prompts produce better outputs.
Templates	A reusable prompt gives consistent results for a repeated task.
Evaluate	Always check accuracy, bias, and relevance.

Vocabulary: tap a card to flip

Tap a card (or press Enter) to reveal its meaning.

Context-setting

Telling the AI what the task is for and who it is for

Constraints

Limits like length, style, or format

Role assignment

Telling the AI who to act as

Iterate

Changing one thing at a time to improve the result

Prompt template

A reusable prompt that produces consistent results

Evaluate

Checking an output for accuracy, bias, and relevance

Knowledge Check

1. 'Context-setting' in a prompt means:

2. Which of these is a prompting technique?

3. 'Iterating' on a prompt means:

4. A prompt template is useful because it:

5. When you evaluate an output, you check:

Lesson Checklist

Used context, constraints, roles, and examples in a prompt
Iterated on a prompt by changing one thing at a time
Built a reusable prompt template and tested it twice
Evaluated an output for accuracy, bias, and relevance