AI Operations - Book Structure

This document describes the logical structure of AI Operations, Book 10 of The Transformed Architecture. The book is organised into seven parts with 31 numbered chapters, moving from the operations in question through the craft of the five operations functions, the operation of AI systems, the operator's own AI, operations in the organisation and the people who run it, to experience and conclusions, followed by the appendices.

Table of contents

Part I: The Operations in Question (Chapters 1–4)

Before the transformation, the pressure, and the operations loop defined (defines O1 Observe to O5 Sustain)

Part II: The Craft (Chapters 5–12)

The craft: observe, observe with AI, diagnose, change, change safely, restore and sustain

Part III: Operating AI Systems (Chapters 13–15)

Operating AI systems: drift and provider change, incidents involving AI and the economics of AI operations

Part IV: The Operator's Own AI (Chapters 16–18)

The operator's own AI: an agent team, the observer principle and defending the operator's AI as Tier 0

Part V: Operations in the Organisation (Chapters 19–24)

Operations in the organisation: resilience as a claim, measurement, operating models, decision rights, regulation and boundaries

Part VI: People and Career (Chapters 25–27)

People and career: operators as people, competences and the future of operations

Part VII: Experience and Conclusions (Chapters 28–31)

Cases, a year in operations, conclusions for two organisations and the verdict

Appendices and Reference

Glossary, Appendix A (code examples: service, runbook, mandate, observer, restore-test and change artefacts), Appendix B (templates and tools), references and about the author

Chapter structure

Part I: The Operations in Question

The part is introduced by part_i_the_operations_in_question.md.

Chapter	File	Title	Description
1	01_introduction.md	Introduction: Operations with AI	The question, the five-part thesis and what would refute it, reading paths for each audience, the two organisations, the usage modes and the degrees of delegation
2	02_what_operations_did_before.md	What Operations Did Before: Runbooks, ITIL, DevOps, SRE and AIOps	A fair history of how operations organised itself, from runbooks, ITIL and ISO/IEC 20000-1 to DevOps, SRE, platform engineering and first-generation AIOps, and the lessons of automation in other domains
3	03_the_pressure_on_operations.md	The Pressure on Operations: Complexity, Dependency, Cost and Machine-Speed Change	Why the old shape strains: complexity, dependency concentration, rising change volume, telemetry and compute cost, scarce skills, regulatory duty and adversaries at the same speed
4	04_the_operations_loop_defined.md	The Operations Loop Defined: Five Functions, the Autonomy Budget and the Observer Principle	The frame: O1 to O5, the autonomy budget with action classes and the Level 4 prohibitions, the observer principle, resilience as a claim with evidence age and the operator's AI as Tier 0

Part II: The Craft

The part is introduced by part_ii_the_craft.md.

Chapter	File	Title	Description
5	05_observe.md	Observe: Telemetry, Observability and Service Level Objectives	O1 Observe: telemetry, service level indicators and objectives, error budgets, service maps, the cost of telemetry and personal data in logs, without AI first
6	06_observe_with_ai.md	Observe with AI: Correlation, Anomaly Detection and Alert Noise	O1 Observe and O2 Diagnose: what AI adds to detecting operational anomalies, its failure modes (suppression, drifting baselines) and how to evaluate it
7	07_diagnose.md	Diagnose: Triage, Root-Cause Analysis and the Incident Assistant	O2 Diagnose: the incident assistant, change correlation, hypothesis ranking, benchmarks read with their conditions, automation bias and the human incident commander
8	08_change.md	Change: Releases, Infrastructure as Code and AI-Authored Change	O3 Change: the change pipeline when agents author much of the change, provenance, review load, policy as code, drift and change records as code
9	09_change_safely.md	Change Safely: Progressive Delivery, Freezes and Blast Radius	O3 Change and O4 Restore: progressive delivery, configuration pushes, freezes, automatic rollback and blast-radius budgets
10	10_restore.md	Restore: Runbooks, Auto-Remediation and Recovery	O4 Restore: runbooks as code, auto-remediation tiers, verification by independent probe and backup and restore as the last line; a constructive debate on whether AI should remediate production on its own
11	11_sustain_capacity_cost_energy.md	Sustain: Capacity, Cost, Energy and Lifecycle	O5 Sustain: capacity planning, FinOps by agents, energy reporting and lifecycle (patching, end-of-life, certificates)
12	12_sustain_problems_and_learning.md	Sustain: Problem Management, Post-Incident Learning and Knowledge	O5 Sustain and O2 Diagnose: problem management, blameless post-incident review, AI-drafted reviews, knowledge hygiene and learning that changes things

Part III: Operating AI Systems

The part is introduced by part_iii_operating_ai_systems.md.

Chapter	File	Title	Description
13	13_operating_ai_systems.md	Operating AI Systems: Drift, Model Updates and Provider Change	O1, O3 and O5: running AI systems as configuration, provider-side model updates, evaluation gates, drift and quality objectives, deployer duties
14	14_incidents_involving_ai.md	Incidents Involving AI: Detection, Triage and Reporting	O2 Diagnose and O4 Restore: what an AI incident is operationally, triage, containment by stop path and reporting clocks under the AI Act, NIS 2, DORA and the GDPR
15	15_the_economics_of_ai_operations.md	The Economics of AI Operations: Tokens, GPUs, Latency and Energy	O5 Sustain and O1 Observe: unit economics of inference, capacity, latency and quality objectives, fallbacks, sovereign capacity, energy and budget guardrails for agents

Part IV: The Operator's Own AI

The part is introduced by part_iv_the_operators_own_ai.md.

Chapter	File	Title	Description
16	16_an_ai_agent_team_for_operations.md	An AI Agent Team for Operations: Skills, Mandates and Containment	O1 to O5: roles for an operations agent team, skills, mandates, containment, memory and tool permissions, evaluation before promotion
17	17_the_observer_principle.md	The Observer Principle: Who Watches the Watcher	O1 Observe and O4 Restore: the central design rule; independence of verification, observer independent of the observed, the stop path outside the control plane; a constructive debate
18	18_defending_the_operators_ai.md	Defending the Operator's AI: Remediation Agents, Credentials and the Control Plane as Tier 0	O3 Change and O4 Restore: threats to remediation agents by mechanism, credentials, the control plane as Tier 0 and the controls that contain them

Part V: Operations in the Organisation

The part is introduced by part_v_operations_in_the_organisation.md.

Chapter	File	Title	Description
19	19_resilience_as_a_claim.md	Resilience as a Claim: Restore Tests, Game Days and Evidence Age	O4 Restore and O5 Sustain: restore tests with measured recovery time and point, game days, evidence age and a resilience register
20	20_measuring_operations.md	Measuring Operations: The Four Key Metrics, SLOs, Toil and the Metrics That Mislead	All functions: the four key metrics, objective attainment, MTTR as a flawed statistic, toil, autonomy-budget consumption and the metrics that mislead
21	21_operating_models.md	Operating Models: Platform Teams, SRE, Managed Services and Sovereign Operations	All functions: in-house SRE, platform teams, managed services, follow-the-sun, sovereign operators and what AI changes in team topology
22	22_decision_rights_and_the_incident_commander.md	Decision Rights, Change Authority and the Incident Commander	O2, O3 and O4: who may approve change, declare an incident, authorise a failover and stop an agent; change authority replacing the change advisory board
23	23_regulation_as_constraint.md	Regulation as Constraint: DORA, NIS 2, CER, the AI Act, the GDPR and the Data Act	All functions: article-level mapping of DORA, NIS 2, the CER Directive, the AI Act, the GDPR and the Data Act to operational controls
24	24_across_boundaries.md	Across Boundaries: Cloud Providers, Managed Providers, Outages and Exit	O1, O4 and O5: concentration risk, shared responsibility in an outage, contracts, registers of information, exit plans and switching

Part VI: People and Career

The part is introduced by part_vi_people_and_career.md.

Chapter	File	Title	Description
25	25_operators_as_people.md	Operators as People: On-Call, Toil, Trust and Wellbeing	All functions: on-call burden, alert fatigue, vigilance and responsibility under automation, working time, co-determination and wellbeing
26	26_competences_and_keeping_the_craft.md	Competences and Keeping the Operator's Craft	All functions: the operator's competences, skill decay and how to prevent it, the early-career route; a constructive debate on deskilling
27	27_the_future_of_operations.md	The Future of Operations: Will Anyone Be on Call in 2036?	All functions: observable, expected and speculation; a constructive debate on whether anyone will be on call in 2036; the author's projections, labelled

Part VII: Experience and Conclusions

The part is introduced by part_vii_experience_and_conclusions.md.

Chapter	File	Title	Description
28	28_case_studies_and_anti_patterns.md	Case Studies and Anti-Patterns	All functions: verified public cases analysed by mechanism and a catalogue of anti-patterns
29	29_a_year_in_operations.md	A Year in Operations: Nordvik Region and Hansa Mutual	All functions: a narrative year for named people in the two organisations, month by month
30	30_conclusions_for_nordvik_and_hansa.md	Nordvik Region and Hansa Mutual: Conclusions for Two Organisations	All functions: a conclusion for each organisation, a twelve-month plan, decisions for the board and the book's own claims as indicators
31	31_conclusion.md	Conclusions: Operations with AI	All functions: the thesis judged against the evidence, what is established, expected and speculative, the author's projections and five things to do on Monday

Appendices and Reference

The appendices are introduced by part_appendices.md.

Item	File	Title	Description
Glossary	32_glossary.md	Glossary	Defines the vocabulary of the book.
Appendix A	33_appendix_a_code_examples.md	Appendix A: Code Examples: Service, Runbook, Mandate, Observer, Restore-Test and Change Artefacts	Fuller versions of the artefacts introduced in the chapters: service level objective record, runbook manifest, mandate manifest, observer manifest, restore-test record, change record, incident record, delegation profile, on-call charter and permission manifest.
Appendix B	appendix_templates_and_tools.md	Appendix B: Templates and Tools	A directory of the templates and tools the chapters provide, with the chapter that introduces each.
References	34_references.md	References and Sources	Canonical bibliography
About the Author	35_about_the_author.md	About the Author	Background and perspective

Companion files

File	Purpose
DISPOSITION.md	The fixed design of the book: premise, thesis, boundaries towards the earlier volumes, red threads, identifiers and a brief for every chapter
BOOK_BLUEPRINT.md	Premise, thesis, audience, the brief of every chapter, the foundations it builds on, the operations functions it concerns and its status
SERIES_FOUNDATIONS.md	The conclusions of Books 1–9 that this book builds on, with the tables of contents of the earlier volumes for cross-references
EXAMPLE_ORGANISATIONS.md	Canonical profiles of Nordvik Region and Hansa Mutual, including their operations profiles
CHAPTER_AUTHORING_GUIDE.md, STYLE_GUIDE.md, AUTHOR_PERSONA.md	Rules for structure, language and voice

Canonical ordering

scripts/toc_definition.py holds the table of contents and regenerates docs/book_index.json, the mkdocs.yml navigation and the chapter list in BOOK_REQUIREMENTS.md. docs/book_index.json and mkdocs.yml must agree (tests/test_book_index.py). Chapter numbers are fixed across the series; later volumes refer back to this book by chapter number.