REST APIs & AI Integration

As an AI Engineer, you do not build giant AI models from scratch. Instead, you use pre-built AI models (like Google Gemini, Claude, or GPT) through APIs and connect them to your own applications.

---

1. What is an API?

An API (Application Programming Interface) is a bridge that lets different software talk to each other.

Simple analogy: Think of a restaurant:

• You = Your Python code (the customer)
• Menu = The API documentation (what you can order)
• Waiter = The API (carries your request to the kitchen)
• Kitchen = The AI model server (does the actual work)
• Food = The API response (the result you get back)

You never go into the kitchen yourself. You just tell the waiter what you want, and the waiter brings back the result.

---

2. GET vs POST Requests

Type	What it Does	Example
GET	Gets data FROM a server	Fetch today's weather
POST	Sends data TO a server	Send a prompt to an AI model

AI APIs always use POST because prompts are long and need to be sent securely inside the request body (not in the URL).

---

3. Making Your First API Call with Python

First, install the requests library:

pip install requests

GET Request Example — Fetching Weather Data:

import requests

url = "https://api.open-meteo.com/v1/forecast"
params = {
    "latitude": 17.38,
    "longitude": 78.47,
    "current_weather": True
}

response = requests.get(url, params=params)
data = response.json()

temp = data["current_weather"]["temperature"]
print(f"Hyderabad temperature: {temp}°C")

---

4. API Keys & Authentication

Most AI APIs need an API key — like a password that proves you are allowed to use the service.

import requests

headers = {
    "Content-Type": "application/json",
    "Authorization": "Bearer YOUR_API_KEY_HERE"
}

• Content-Type tells the server you are sending JSON data.
• Authorization proves your identity using your API key.

⚠️ Never share your API key publicly! Never put it directly in your code. We will learn a safer way in the next lesson.

---

5. Keeping API Keys Safe with dotenv

Create a .env file in your project folder:

GEMINI_API_KEY=your-secret-key-here

Then load it in Python:

import os
from dotenv import load_dotenv

load_dotenv()  # Load the .env file

api_key = os.getenv("GEMINI_API_KEY")
print("Key loaded!" if api_key else "Key not found!")

Install dotenv:

pip install python-dotenv

Why this matters: In real companies, API keys are never written directly in code. They are stored in .env files (which are added to .gitignore so they never get pushed to GitHub).

---

6. Calling an AI Model API (Google Gemini)

import requests
import os
from dotenv import load_dotenv

load_dotenv()
api_key = os.getenv("GEMINI_API_KEY")

url = f"https://generativelanguage.googleapis.com/v1beta/models/gemini-1.5-flash:generateContent?key={api_key}"

payload = {
    "contents": [
        {
            "parts": [
                {"text": "Explain what Machine Learning is in 2 simple sentences."}
            ]
        }
    ]
}

response = requests.post(url, json=payload)
data = response.json()

# Extract the AI's answer
answer = data["candidates"][0]["content"]["parts"][0]["text"]
print("AI says:", answer)

---

Summary

• API = A bridge that lets your Python code talk to AI model servers.
• GET = Fetch data from a server. POST = Send data to a server.
• AI APIs use POST requests with JSON payloads.
• Always keep API keys in .env files — never hardcode them.
• You can call any AI model (Gemini, Claude, GPT) using Python's requests library.