World Congress 2022 • Jun 15, 2022

Build your backend using FastAPI

Ashmi Banerjee

Tired of heavy Django boilerplate and manual endpoint testing? Build your next machine learning backend with FastAPI to get out-of-the-box data validation and automatic interactive documentation.

Pause
Mute Enter Fullscreen
#1 about 3 min

Prerequisites for building backends with FastAPI

Basic knowledge of Python, machine learning, and web development provides the foundation for building API backends.

#2 about 3 min

Traditional versus modern web architecture for rendering pages

Modern web architecture pairs content delivery networks for static content with API servers returning JSON for dynamic components.

#3 about 7 min

Comparing Python backend frameworks for web development

Django offers full-stack robustness while Flask and FastAPI provide lightweight alternatives with reduced boilerplate for API-focused applications.

#4 about 2 min

Key features and advantages of using FastAPI

FastAPI provides automatic interactive documentation, asynchronous IO support, and rapid development capabilities for REST endpoints.

#5 about 6 min

Serving a machine learning image classifier via API

Wrapping a pre-trained machine learning model inside an API makes it broadly accessible across applications without local downloads.

#6 about 2 min

Interactive demonstration of the image classifier API

An interactive Swagger interface allows testing the endpoints directly using image URLs to evaluate real-time model predictions.

#7 about 3 min

Testing, containerization, and deployment for API applications

Moving an application to production requires thorough testing, Docker containerization, and safe deployment to a commercial cloud service.

#8 about 2 min

Model serving options and asynchronous function requirements

Evaluating tools like TensorFlow Serve offers alternatives to FastAPI while highlighting specific rules for asynchronous function declarations.

Matching moments

1:42 min

Introduction to the fast API web framework

Sebastián Ramírez · World Congress 2022

1:50 min

Introduction to building APIs with Flask

Teri Eyenike · LIVE

2:18 min

Exploring the tiered architecture of modern machine learning stacks

Kris Howard · LIVE

3:33 min

Connecting frontends via a FastAPI proxy backend layer

Saoussen Chaabnia Saoussen Chaabnia · Europe 2026 Virtual

3:10 min

Code and architecture for building the discount application

Bobur Umurzokov · LIVE

3:54 min

Deploying and testing the AI inference service

Daniel Brintzinger Daniel Brintzinger · World Congress 2025