"NIST AI RMF in Engineering: Turning Risk Management into Shipping Controls"
AI teams rarely fail because they lack principles; they fail because principles never become enforceable delivery controls. This book is written for experienced engineers, technical leads, platform owners, security architects, ML practitioners, and governance stakeholders who need to turn the NIST AI Risk Management Framework into concrete release gates, runtime safeguards, and operational decision systems. It treats AI risk management not as policy theater, but as a shipping discipline embedded in real engineering work.
Across the book, readers learn how to translate GOVERN, MAP, MEASURE, and MANAGE into system inventories, accountability models, context mapping, TEVV pipelines, evidence standards, launch criteria, monitoring architectures, incident playbooks, and change-control mechanisms. The emphasis is practical and deeply technical: risk tolerance becomes release criteria, trustworthiness becomes measurable engineering objectives, and residual risk becomes something teams can assess, document, and operate with under production pressure. Special attention is given to third-party components, supply-chain dependencies, and generative AI extensions.
The book assumes familiarity with software delivery, MLOps, and production operations, and it rewards readers who already work at the boundary between engineering and governance. Its distinguishing strength is structural clarity: each chapter gives the NIST AI RMF a concrete operational home inside the lifecycle of building, releasing, and running
