From Prompt Engineering to AI Applications

Business Scenario

Welcome!

Today is your final day as a ML Engineer at FusionTech Consulting 

In the previous lab, you explored different Generative AI models, designed effective prompts, experimented with various prompting techniques, and learned how to control AI-generated responses

Your responsibility is to build a simple AI-powered application using programming and an API. Instead of typing prompts into a browser, your program will send prompts directly to the AI model and receive responses automatically.

These concepts will help you:

• Connect Python applications with AI models

• Authenticate AI services using an API Key

• Send prompts programmatically

• Build AI-powered software applications

Pre-Lab Preparation

Topic: From Prompt Engineering to AI Applications

1) Large Language Models (LLM)

2) API Key

3) Programming Integration

Git Pull

Task 1: Connect Python to Google Gemini API

git pull origin branchName

1

Open Web Browser and visit this website

https://aistudio.google.com

Sign in using your Google Account.

2

Click Create API Key

3

Click on Create Key

4

API Key will be generated

5

AIzaSyAxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 

A valid key looks like this :

Open Jupyter Notebook 

6

ai_application.ipynb

Create a new Notebook and name it as :

7

Install the Google Gemini Python SDK

8

import sys

!{sys.executable} -m pip install -U google-genai

Restart the Jupyter Kernel

9

Verify that the SDK has been installed successfully

10

from google import genai

print("Google GenAI SDK Installed Successfully")

Create a new code cell

11

from google import genai

API_KEY = "PASTE_YOUR_API_KEY_HERE"

MODEL_NAME = "gemini-3.5-flash"

client = genai.Client(
 api_key=API_KEY
)

prompt = """
Explain Artificial Intelligence in simple language suitable for beginners.
"""

response = client.models.generate_content(
 model=MODEL_NAME,
 contents=prompt
)

print("\n========== AI Response ==========\n")
print(response.text)

If Program displays error :

12

ClientError: 404 NOT_FOUND 

Execute the following code to view the available Gemini models.

from google import genai

client = genai.Client(api_key=API_KEY)

for model in client.models.list():
 print(model.name)

Replace :

MODEL_NAME =

Key Insertion :

13

Replace

With the API Key you created

Example :

API_KEY = "PASTE_YOUR_API_KEY_HERE"
API_KEY = "AIzaSyAxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"

Run the Program & save the Notebook

14

Locate the following prompt

15

Replace it with : 

prompt = """
Explain Artificial Intelligence in simple language suitable for beginners.
"""
prompt = """
Explain Prompt Engineering in less than 100 words using bullet points.
"""

Run the Program & Observe the output

16

Task 2: Build an AI Topic Explainer

Open the file created in Task 1

1

Locate and Delete the following prompt :

2

prompt = """
Explain Prompt Engineering in less than 100 words using bullet points.
"""

Create a variable to accept user input

3

Type the following statement immediately below the client object

topic = input("Enter Topic : ")

Create a reusable prompt template

4

prompt = f"""
You are an experienced Technical Trainer.

Task:
Explain the given topic.

Context:
The explanation is for first-year engineering students who are beginners.

Constraints:
• Maximum 150 words
• Use simple language
• Include one real-world example
• Avoid difficult technical terms

Output Format:
Definition
Key Features
Real-World Example
Conclusion

Topic:
{topic}
"""

Run the Program, Input the Topic and Save the file

5

Enter Topic : Artificial Intelligence

Try modifying the constraints and observe the response

6

Try modifying output format section

7

Overview

Advantages

Applications

Summary

Task 3: Control AI Responses

1

Open the file and replace the prompt with :

prompt = f"""
You are an experienced Technical Trainer.

Task:
Explain the given topic.

Context:
The explanation is for first-year engineering students.

Constraints:
• Maximum 150 words
• Use simple language

Output Format:
Paragraph

Topic:
{topic}

"""

2

Run the Program

Replace the Output Format

3

Enter Topic : Artificial Intelligence

Output Format:

Definition

Key Features

Advantages

Applications

Conclusion

Try modifying Output Formats

4

Modify Input Section

5

topic = input("Enter Topic : ")

topic = input("Enter Any Technology Topic : ")

Replace with

Task 4: Build your AI Assistant

1

Open the file and replace the prompt with :

prompt = f"""
You are an AI Email Writing Assistant.

Task:
Write a professional email based on the user's request.

Context:
The email should be suitable for workplace communication.

Constraints:
• Use a professional tone.
• Maximum 150 words.
• Generate an email with a Subject and Body.
• Do not use complicated vocabulary.

Output Format:
Subject:
Email Body:
User Request:
{topic}
"""

Locate the following

2

topic = input("Enter Any Technology Topic : ")

topic = input("Enter your Email Request : ")

Replace with

Run the Program

3

Enter Topic : Write an email requesting two days leave due to illness.

Convert the application into an AI SQL Query Generator

4

Replace the prompt with

prompt = f"""
You are an SQL Expert.

Task:
Generate an SQL query.

Constraints:
• Generate only SQL code.
• Do not provide explanation.

User Requirement:
{topic}
"""

Locate the following

5

topic = input("Enter your Email Request : ")

topic = input("Enter SQL Requirement : ")

Replace with

Enter Topic : Display all employees whose salary is greater than 50000

Run the Program

6

Modify the Topic & Run the Program

7

Display the top 5 highest selling products.

 

Great job!

You have successfully completed the final lab and made your own AI application using Programming and API Keys

Checkpoint

   Git Push

git push origin branchName

Prompt Engg 4

By Content ITV

Prompt Engg 4

  • 26

More from Content ITV