Building Generative AI Services with FastAPI

Chapter 1. Introduction

Chapter Goals

In this chapter, you will learn about:

• Generative AI (GenAI) and its use cases

• Why GenAI services will power future applications

• The barriers to broader adoption of AI services

• How to build a GenAI service

• Why FastAPI is a good choice for building GenAI services

• The overview of the capstone project

By the end of this chapter, you should be able to identify the role of GenAI within the roadmap of your own applications and its associated challenges.

What Is Generative AI?

Generative AI is a subset of machine learning that focuses on creating new content using a model trained on a dataset. The trained model, which is a mathematical model representing patterns and distributions in the training data, can produce new data that is similar to the training dataset.

To illustrate these concepts, imagine training a model on a dataset containing images of butterflies. The model learns the complex relationships between pixels in images of butterflies. Once trained, you can sample from the model to create novel images of butterflies that didn’t exist in the original dataset. These images will contain similarities with the original images of butterflies yet remain different.

Note

Using a trained generative model to create new content based on patterns learned from the training data is known as inference.

Figure 1-1 shows the full process.

Figure 1-1. A generative model trained to create new photos of butterflies

Since we don’t want to ...