Agentic Reliability Engineering

Part III. Making Systems Reason

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 3 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].

Four chapters. Part II showed agents acting; Part III opens up how they think. Decision intelligence, reasoning patterns, governance under autonomy, and simulation as the substrate for trust.

Part II treated the Reasoning Plane as a structured black box. Signals went in, candidate actions came out, the Governance Plane evaluated against policy, the Execution Plane ran the chosen action under bounded scope. Across five chapters and four workflows, that black box was sufficient. The architecture worked because the shape of reasoning was specified (structured artefacts, confidence-bearing claims, evidence-traceable outputs) even when the internals were not. A reader could follow incident response, delivery, chaos, and pre-incident intelligence without ever needing to know how the Reasoning Plane actually arrived at its outputs. That was deliberate. Part II’s job was to install the architectural shape; opening the box would have made every workflow chapter twice as long ...