AI Prompt

Engineering
QuickStart Guide
with Tina Huang
Why This Matters
AI models like ChatGPT are incredibly powerful, but the

quality of their output depends largely on the quality of your

prompt. Just like a good teacher needs clear instructions to

help students succeed, AI needs well-structured prompts to

deliver its best work. This guide will help you master prompt

engineering to get consistently great results.

AI Prompt Engineering QuickStart Guide 2

Framework
Breakdown
1. Set the Stage (Context)

Tell the AI who 
 Define the 
 Specify any

it's talking to expertise level constraints or
needed limitations

Bad Prompt → "Analyze this dataset"

↓
Good Prompt → "Analyze sales trends in this CSV"
↓
Excellent Prompt →
"You're a data scientist analyzing our quarterly sales
data. Identify key revenue patterns, create
visualizations, and highlight anomalies. Focus on
customer segments and regional variations."

AI Prompt Engineering QuickStart Guide 3

2. Define the Task (Action)

State exactly 
 Include specific Break complex 


what you want requirements tasks into steps

Bad Prompt → "Help me learn Python"

↓
Good Prompt → "Teach me Python data structures"
↓
Excellent Prompt →
"You're a senior Python instructor. Create a step-by-
step learning plan for data structures, including
Practice exercises for lists and dictionarie
Real-world examples using panda
Common pitfalls to avoid"

AI Prompt Engineering QuickStart Guide 4

3. Specify the Output (Format)

Define the Set the 
 Clarify the 


structure length tone and style

Bad Prompt → "Make a content plan"

↓
Good Prompt → "Create a formatted content plan with specific sections"
↓
Excellent Prompt →
"Present the plan in a table format with columns for
platform, content type, and key message. Include
engagement hooks for each piece. Keep individual
posts under 200 characters."

AI Prompt Engineering QuickStart Guide 5

Power Tips for 

Better Results
Tip 1
Use Role Assignment
When you assign a specific role to AI, you tap into specialized knowledge and perspectives. Think of
it like hiring an expert consultant – the more specific you are about their expertise and experience,
the more tailored their advice will be. This technique is particularly powerful when you need
industry-specific insights or want to ensure the output matches a certain professional standard.

Bad Prompt → "Debug this code"

↓
Good Prompt → "Debug this Python machine learning model"
↓
Excellent Prompt →
"As a senior ML engineer specializing in neural networks
Review this PyTorch implementatio
Identify performance bottleneck
Suggest optimization strategies for training speed"

AI Prompt Engineering QuickStart Guide 6

Tip 2
Add Specific Constraints
Constraints aren't limitations – they're clarity boosters. By setting clear parameters, you help the AI
understand exactly what success looks like. This is like giving a designer exact specifications
instead of asking them to "make it look good." The more specific your constraints, the more
targeted and useful the output will be.

Bad Prompt → "Analyze this dataset"

↓
Good Prompt → "Analyze customer churn patterns"
↓
Excellent Prompt →
"Analyze our customer churn dataset with these
parameters
Focus on users inactive > 30 day
Compare behavioral patterns pre-chur
Include statistical significance test
Output visualizations in seaborn"

AI Prompt Engineering QuickStart Guide 7

Tip 3
Request Examples and Variations
Getting multiple versions of the same content helps you explore different approaches and find the
perfect fit. This technique is particularly valuable when you're not sure which style will work best, or
when you need to compare different approaches. It's like having a creative team pitch multiple
concepts – you can mix and match the best elements from each version.

Bad Prompt → "How do I implement this?"

↓
Good Prompt → "Show different ways to implement this algorithm"
↓
Excellent Prompt →
"Provide 2 implementations of this sorting algorithm
Optimized for memory efficienc
Optimized for processing speed Include time
complexity analysis for each"

AI Prompt Engineering QuickStart Guide 8

Troubleshooting
Guide
If your AI response isn't quite right:

1 Add more specific 


context about your
audience and goals 2 Break your request into
smaller, clearer steps

3 Include examples of
what good looks like
4 Ask for variations and
iterate on the best one

AI Prompt Engineering QuickStart Guide 9

Additional Resources

Breeze AI

Experience the transformative power of HubSpot’s

AI that powers the entire customer platform.
Breeze AI delivers ease, speed, and unification
across your business, giving customer-facing
teams all the AI tools needed to boost productivity,
scale growth, and unlock actionable insights.

Tina Huang

Learn more from data scientist Tina Huang. With over

680,000 YouTube subscribers, she decodes complex AI
concepts into practical, efficient strategies that deliver
real results. Her step-by-step tutorials and innovative
approaches help professionals maximize AI's potential
while minimizing effort. Discover how to transform your
AI interactions from basic to brilliant.