Agentic Reliability Engineering

Part I. Why Reliability Must Evolve

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 Part 1 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].

The structural limit of human-paced reliability, the observability that any agentic system depends on, and the resilience loop that turns failure into a signal the system learns from. This is what the next three chapters will cover.

The Prologue made the case for why the discipline this book describes is arriving now. Part I makes the case for why the disciplines that came before it cannot absorb the work that is coming. Three chapters do that work, in a deliberate order, and each of them installs something the rest of the book will refer back to.

Chapter 1 is the diagnosis chapter. It opens with a senior on-call engineer at 03:00, doing everything the role asks for, on a system that has grown beyond what the role can hold. The chapter argues that the ceiling SRE has been pressing against for the last few years is not a tooling ceiling or a process ceiling or a hiring ceiling. It is a cognitive ceiling, and no amount of additional automation removes it because automation ...