Cover of Foundations of Agentic AI Systems, Volume II: Agents, Operations, and Governance, in Production, by Fatih Nayebi, on a wine-red ground

Volume II · October 2026

Foundations of Agentic AI Systems, Volume II

Agents, Operations, and Governance, in Production

Every grant of authority, earned on evidence named in advance.

Volume II builds the refund agent of Volume I for production. It gives the agent a loop that ends every run in a known way, a gate inside the call that pays, an independent verifier, and a release suite that earns shadow mode: every decision made and recorded, no money moved.

Hardcover on Amazon Companion code

About the book

Then come the operations a deployment needs: owned data and its freshness, a canary with a rehearsed rollback, error budgets, scoped credentials. Governance is treated as engineering: an outcome audit with a test, an explanation for every denial, a rollout plan that says how each person’s work changes. The capstone sets out the launch packet and its gate, and the closing chapters carry the same decision packet into three other domains.

The book is plain about what it has run and what it has only designed. The companion code holds the records through shadow mode and the designs after it, and the text says which is which.

Builds on Volume I, Models, Learning, and Decisions, from First Principles; a bridge at the front recaps what it uses.

Contents

Part V · Agentic Systems Architecture

  1. 22 The Agentic Loop
  2. 23 Perception and Tools
  3. 24 Reasoning and Validated Planning
  4. 25 Action, Guardrails, and Policies
  5. 26 Memory and Retrieval
  6. 27 Multi-Agent Systems and Independent Verification
  7. 28 Agent Frameworks and Orchestration
  8. 29 Agent Evaluation and Release Evidence

Part VI · Engineering and Operations

  1. 30 Data Systems for Agentic AI
  2. 31 Batch and Streaming Pipelines
  3. 32 Deployment and Rollback
  4. 33 MLOps and ML Engineering
  5. 34 Observability and Reliability
  6. 35 Security and Privacy

Part VII · Governance and Adoption

  1. 36 Outcome Audits and Decision Explanations
  2. 37 Human Oversight and Risk Management
  3. 38 Organizational Adoption

Part VIII · Capstone and Futures

  1. 39 Capstone: Design, Build, Deploy, Operate
  2. 40 Domain-Agnostic Transfer Cases
  3. 41 Future Directions and System Renewal
Fatih Nayebi

About the author

Fatih Nayebi

Fatih Nayebi is Assistant Professor in Information Systems at McGill University, where since 2019 he has taught the five graduate courses these volumes hold: Enterprise Data Science, Machine Learning in Production, Introduction to AI and Deep Learning, Applications and Architectures of Deep Learning, and Designing and Building Agentic AI Systems. As founder and CEO of OODARIS AI, he builds real-world, production multi-agent systems that make autonomous decisions, combining machine learning, ML engineering, deep reinforcement learning, operations research and software engineering, and that act only within the authority they have earned.

More about Fatih OODARIS AI

Also in the series

All books

Autonomy should rise only as evidence rises. · Published by Gradient Divergence, Montréal.