Exercise 5: Role-Based Scenario Practice

Objective

Apply the full prompt workflow — Frame, Aim, Guide, Refine, Reuse — to a realistic scenario from your own role.

Background

This is the capstone exercise of the course. It asks you to take a real scenario from your actual job and walk through the complete process: building a strong initial prompt, evaluating the output, refining it with specific instructions, and converting the best version into a reusable template.

This is not a test. It is a practice run for what good AI use looks like in your real context.

Task

Step 1: Choose a scenario

Pick a real, near-term task from your actual work that you would typically use AI to help with. It should be something that has a real audience, a real purpose, and a real output type (email, memo, document, summary, plan, etc.).

If you prefer not to use real work content, choose a scenario from the role examples below.

Role scenarios (choose one if not using your own)

Product manager:
You need to write a one-page brief summarizing a proposed feature for an upcoming sprint planning meeting. Stakeholders include engineering leads and the VP of Product. The feature in question allows users to set recurring reminders inside your project management tool.

Marketer:
You are drafting a re-engagement email for customers who have not logged in for 60 days. Your product is a document collaboration tool. The tone should nudge without being pushy.

Support / documentation:
You are writing a help article explaining how to set up two-factor authentication. Audience: regular users who are not technical. Many of them will be reading this because they got locked out.

Operations coordinator:
You are writing a post-mortem summary after a vendor failed to deliver on a key deadline. Audience: internal leadership. The summary should cover what happened, what the impact was, and what you are changing going forward.

Developer:
You are writing a brief technical explainer on why the team chose PostgreSQL over MongoDB for a new feature's data layer. Audience: non-technical stakeholders in a product review meeting.

Step 2: Build the initial prompt

Using the 5-step framework from the capstone, write your full prompt:

• Frame: What is this and why does it matter?
• Aim: Who is it for? What outcome do you want?
• Guide: Tone, format, constraints?
• (Refine and Reuse come after you see the output)

Step 3: Submit and evaluate

Submit your prompt and read the output critically using the five checks from Lesson 8:

1. Does it accomplish the goal?
2. Is it organized for the reader?
3. Is the tone right?
4. Is there anything to cut?
5. Is there anything missing?

Step 4: Write one or two targeted refinement instructions

Based on your evaluation, write one or two specific follow-up instructions. Submit them. Evaluate the improved output.

Step 5: Build the reusable template

If this task type will recur (even occasionally), convert your refined prompt into a template. Mark the variable elements as placeholders. Keep the fixed structure intact.

Success criteria

By the end of this exercise, you should have:

• A well-framed initial prompt built on the 5-step framework
• At least one specific follow-up refinement instruction
• A final output you would be comfortable using or sharing at work
• A reusable template for future instances of this task (if applicable)

Optional reflection question

Looking back at the prompts you wrote at the start of this course and the ones you are writing now — what is the most significant change in how you approach an AI request?