Building Live Voice Agents

Chapter 6. Designing and Building Agents

A Note for Early Release Readers

With Early Release ebooks, you get books in their earliest form—the author’s raw and unedited content as they write—so you can take advantage of these technologies long before the official release of these titles.

This will be Chapter 6 of the final book. Please note that the GitHub repo will be made active later on.

If you’d like to be actively involved in reviewing and commenting on this draft, please reach out to the editor at [e-mail].

While building an agent execution loop from the ground up is a fantastic way to learn, production systems cannot rely on custom-built orchestration for each project. Developers typically need a dedicated framework to manage state, route tool calls, and execute the reasoning loop so they can focus on the application’s domain logic.

This chapter moves from framework theory to concrete implementation using the Google Agent Development Kit (ADK) as an example. We utilize ADK because its primitives map directly to the system boundaries required for enterprise applications:

Modular agent design

Agents are constructed as discrete, composable building blocks, allowing them to act as simple workflow controllers or complex reasoning engines.

Easy tool creation

Tool creation is cleanly abstracted into standard Python functions, giving agents the ability to interact with the outside world easily.

Comprehensive session management

Short-term and long-term memory are handled ...