8 Database design in the age of generative AI

 

In this chapter

  • You explore how to use generative AI to speed database design.
  • You learn how to fine-tune the prompts to get precisely what you want from generative AI tools.
  • You go through the full pipeline to design and implement a database for a spaceship manufacturer with the help of generative AI tools.
What you need to know

This chapter is different from others in a major way. You are unlikely to replicate the demonstrated interactions with large language models (LLMs) due to their probabilistic nature and continuous changes. When we wrote this chapter, we used ChatGPT 4o. That said, ChatGPT may get your design exactly right or wrong in different ways even if you use the prompts demonstrated in this chapter.

The database design covered in this chapter (it’s finally complete), using tools commonly used by practitioners, such as dbdiagram.io and MySQL Workbench, can be found in the GitHub repository (https://github.com/Neo-Hao/grokking-relational-database-design). Navigate to the chapter_08 folder and follow the instructions in the README.md file to load the database design into corresponding tools.

Also, you can find the full-text prompts for ChatGPT and SQL scripts corresponding to the final database design for different relational database management systems (RDBMS), including MySQL, MariaDB, PostgreSQL, SQLite, SQL Server, and Oracle.

Overview

Requirements analysis

Entities and attributes

Entities

Attributes

Communicate your design

Relationships among entities

Minimum cardinalities and foreign keys

Communicate your design

Implementation

Constraints