Chapter 3 of 3

Advanced Prompting Patterns

1. Structured Output

Ask the AI to return data in a specific format — JSON, markdown tables, CSV. This is essential for integrating AI into applications.

Structured output prompttext

Extract the following information from this job posting and return it as JSON:
- Job title
- Company name
- Location
- Salary range
- Required skills (as an array)

Job posting: "Senior React Developer at TechCorp, Remote (US), $120k-$160k. Must know React, TypeScript, Node.js, and PostgreSQL."

→ {
  "title": "Senior React Developer",
  "company": "TechCorp",
  "location": "Remote (US)",
  "salary": "$120k-$160k",
  "skills": ["React", "TypeScript", "Node.js", "PostgreSQL"]
}

2. System Prompts

System prompts set the AI's behavior for an entire conversation. They define the persona, rules, and constraints that apply to every response.

System prompt exampletext

SYSTEM: You are a friendly, encouraging coding tutor for beginners.
Rules:
- Always explain concepts with real-world analogies
- Include code examples in every response
- If the user makes an error, explain why gently
- Never use jargon without defining it first
- Keep responses under 300 words

3. Prompt Chaining

Break complex tasks into sequential steps, where the output of one prompt becomes the input for the next. This produces dramatically better results than trying to do everything in a single prompt.

Prompt Chain: Blog Post Generator

1️⃣

Step 1

Generate 5 blog post title ideas on 'React Hooks'

2️⃣

Step 2

Pick the best title and create an outline

3️⃣

Step 3

Write each section based on the outline

4️⃣

Step 4

Review and refine for tone and accuracy

Key Takeaway

Prompt engineering is a superpower. The techniques you've learned — zero-shot, few-shot, chain-of-thought, structured output, and prompt chaining — will make you 10x more effective with any AI tool.

← Previous

Core Prompting Techniques