GET and POST requests for AI

Deploying AI into Production with FastAPI

Matt Eckerle

Software and Data Engineering Leader

Our instructor

Matt Eckerle

Software and Data Engineering Leader
Enterprise Data Manager at Inari
Using FastAPI for ML since 2019

Course overview

FastAPI fundamentals
Request handling and integration
Input validation and security
Create and maintain production-ready APIs

Logo of FastAPI

Before we start

✓ Basic python: functions, classes, modules, data structures

✓ HTTP & REST API concepts

✓ Using the FastAPI framework

Handling GET and POST requests
Using Pydantic models

✓ Machine learning basics: create, save, predict

The word "basics" written in chalk

GET requests explained

Used to retrieve data
Path parameter - information in URL
Doesn't change server state

GET https://example.com/?item_id=1

A diagram to understand GET requests

Implementing GET with path parameters

from fastapi import FastAPI

app = FastAPI()

@app.get("/item/{item_id}")
async def read_item(item_id: int):
    return {"item_id": item_id}

`@app.get` decorator defines endpoint.
{item_id} is a path parameter.
Type hint (int) for auto validation.

POST requests explained

Used to send data to server
Data in request body (usually JSON)
Can change server state

A restaurant order

POST https://example.com

A diagram to understand POST requests

Implementing POST with JSON data

from fastapi import FastAPI
from pydantic import BaseModel

app = FastAPI()
db = {}

class Item(BaseModel):
    name: str
    price: float

@app.post("/items", status_code=201)

def create_item(item: Item):
      db[item.name] = item.model_dump()
    return {"message": 
            f"Created {item.name}"}

Pydantic model defines data structure
@app.post for POST endpoint
Automatic JSON parsing and validation

HTTP status codes

200 OK: Default for success
404 Not Found: Resource not found
201 Success: Object created

An HTTP 200 status code indicating successful request

Raising HTTP exceptions

from fastapi import FastAPI, HTTPException
app = FastAPI()

@app.get("/item/{item_id}")
async def read_item(item_id: int):
    if item_id == 42:
        raise HTTPException(status_code=404, detail="Item not found")
    return {"item_id": item_id}

@app.post("/items")
async def create_item(item: Item):
    # Simulating item creation
    return {"message": f"Created {item.name}"}, 201

Let's practice!

Deploying AI into Production with FastAPI