Unit Title: Designing Prompt Workflows and Intelligent Input Sequences
Level: Advanced
Duration: 120–150 minutes (may be completed in 2–3 sessions)
🎯 Learning Objectives
By the end of this week, you should be able to:
- Understand prompt chaining and multi-turn reasoning.
- Build prompt systems for repeated tasks, workflows, or toolkits.
- Create personality-aligned or role-specific prompts for custom AI interactions.
- Combine clarity, iteration, and conditional logic in real-world cases.
🧭 Lesson Flow
| Segment | Duration | Format |
|---|---|---|
| 1. Prompt Chains: What and Why | 20 min | Definitions + Process Mapping |
| 2. Systems Thinking with AI | 25 min | Modular Templates + Reuse |
| 3. Role/Persona Prompts | 25 min | Use Cases + Examples |
| 4. Structured Outputs and Consistency | 20 min | Format Prompts + Tools |
| 5. Exercises + Knowledge Check | 40–60 min | Prompt Labs + Refinement |
🧑🏫 1. Prompt Chains: What and Why
📖 Teaching Script:
When a single prompt isn’t enough, we build a chain — a series of inputs where the output of one becomes the input of the next.
Prompt chaining is powerful because it:
- Breaks down complexity into stages
- Builds memory/context progressively
- Encourages deeper and more accurate results
🧪 Real Examples of Prompt Chains:
- Research Assistant Chain
- Prompt 1: “Summarise this article in 5 bullets.”
- Prompt 2: “List 3 questions raised by this summary.”
- Prompt 3: “Answer those questions with citations.”
- Writing Chain
- Step 1: “Generate a headline and subheading for a blog about AI in healthcare.”
- Step 2: “Write a 3-paragraph blog post using that headline.”
- Step 3: “Now revise it to match a friendly tone and add a call to action.”
- Design Brainstorm Chain
- Ask for 10 ideas → Filter by feasibility → Expand on top 3 ideas
✍️ Diagram:
[Prompt 1] ➜ Output A
[Prompt 2: Based on A] ➜ Output B
[Prompt 3: Based on B] ➜ Final Result
🧩 2. Systems Thinking with Prompts
📘 What Is a Prompt System?
A prompt system is a repeatable, reusable structure — like a checklist or framework — that you use regularly to get consistent AI results.
🧪 Three Examples of Prompt Systems:
- Lesson Builder System
- Structure: [Subject] + [Learning Outcomes] + [Activities]
- Prompt: “Create a lesson plan on [X] with objectives, intro, and one interactive task.”
- Analysis System
- Structure: Describe + Compare + Evaluate
- Prompt: “Analyse [issue] by giving an overview, comparing two sides, and offering one reasoned opinion.”
- Social Media Generator
- Prompt 1: “Write a tweet about [X] with a question hook.”
- Prompt 2: “Now convert that into a 4-slide Instagram carousel.”
✏️ Activity:
Choose one task (e.g., writing, planning, researching)
→ Build a 3-part system using prompts
→ Test it and revise
🧑🎓 3. Role/Persona Prompting (Custom Agent Building)
📘 What Are Role Prompts?
You can assign a role or persona to the AI to shape its voice, style, and expertise.
Think of it as telling the AI who it is before you ask what it should do.
🧪 Three Role Prompt Examples:
- “Act as a policy advisor. I’m preparing for a panel on AI regulation. Help me outline 5 key risks and 3 proposed responses.”
- “You are a UX designer. Critique the following app layout for accessibility and clarity.”
- “Assume you’re a spiritual mentor. Write a 300-word reflection on digital silence and inner peace.”
👥 Persona Alignment Prompts:
Try starting prompts like:
- “As a career coach…”
- “Write from the view of an environmental journalist…”
- “Imagine you are a curious 12-year-old asking about space…”
✏️ Activity:
Write 2 prompts using role-based framing: one professional, one creative.
🧾 4. Structured Outputs and Consistency
📘 Use Structured Prompts to Control Format:
| Task | Prompt Structure |
|---|---|
| Product comparison | “Compare [X] and [Y] using a table with pros, cons, and price.” |
| Case summary | “Summarise this in 3 parts: Situation, Problem, Recommendation.” |
| Interview preparation | “Create 5 questions and ideal sample answers for [job role].” |
🧪 Format Examples:
- Table Format Prompt: “List 3 AI tools with columns: Tool Name | Use Case | Free or Paid | Limitation”
- Bullet Structure Prompt: “Summarise 5 benefits of prompt engineering using bullet points and 1 real-world example per bullet.”
- Checklist Prompt: “Create a checklist for ethical use of generative AI in academic writing.”
🧠 Tip:
Always specify the format you want (table, list, summary, script, chart, timeline, etc.)
🧪 5. Exercises + Knowledge Check
✅ Exercise 1: Chain a Prompt
Design a 3-step prompt chain for:
- Writing a policy brief
- Generating 3 visual concepts
- Planning a 1-week digital detox challenge
✅ Exercise 2: Build a Prompt System
Choose a daily/weekly task in your life. Create a reusable system of 2–3 structured prompts to support it (e.g., productivity, planning, research).
✅ Exercise 3: Role Prompt Lab
Craft a role-based prompt for:
- A lawyer giving advice to a small business
- A nutritionist giving feedback on a meal plan
- A coach helping a student set goals
Write both the prompt and expected output format.
🧠 Knowledge Check (10 Questions)
- What is a prompt chain?
- Give one reason why chaining prompts is useful.
- What’s the difference between a chain and a system?
- Name 2 examples of role-based prompting.
- How does assigning a role change the output?
- What’s the benefit of specifying output format?
- How can you prompt for a table?
- What’s a checklist prompt useful for?
- Why is systematisation powerful in AI use?
- Create a 3-part prompt sequence for a personal project.
📝 Wrap-Up Assignment (Optional)
Title: “My Prompt Playbook: From One-Liners to Systems”
Write:
- One full prompt chain (at least 3 steps)
- One prompt system (with format notes)
- One persona-based prompt and result
Conclude with:
- What changed in your thinking about AI since learning these advanced tools?
📦 End-of-Week Deliverables
- ✅ 3 prompt chains built and tested
- ✅ 1 structured system prompt created
- ✅ 2 role/voice prompts developed
- ✅ Knowledge check completed
- ✅ Reflection on prompt mastery