Part 4. Maintenance

 

Congratulations! You’ve built and tested your AI assistant, and now you’re ready to deploy it. This section of the book helps you keep your assistant in good working order as it matures. It’s exciting to build a new assistant. But your assistant will spend much more time in production than you spent in that initial build phase. It’s important to have good maintenance hygiene and processes to keep your assistant working well.

Chapter 9 shows you how to keep track of the software assets that make up your assistant. It introduces source code management and shows how to manage multiple versions of your assistant at the same time. You will have different needs for your development, test, and production environments. This chapter shows you how to balance all those needs.

Chapter 10 teaches you how to analyze and improve your assistant. No assistant will be perfect, especially not on the first release. This chapter shows you how to find the problem spots in your assistant that have the biggest impact on your success metric. It introduces several methodologies that you can use not only to find these problems, but to fix them as well.

After reading this part, you’ll be able to keep your assistant running smoothly in production. Getting to production is not the end; it’s only the beginning.