AI Governance

What changes when an AI assistant can take action? Follow a fictional enterprise team as it works through permissions, prompt injection, testing, monitoring, and governance, using practical diagrams, published cases, and exercises to design and operate agents with clear boundaries

Business and board leaders who need to understand cyber risk; solution and enterprise architects designing services; IT and security professionals working across teams; consultants advising customers; and service owners, people managers, and communications professionals with responsibilities during disruption.

AI가 자신의 과거를 기억하고, 경험을 통해 변화하며, 어제의 자신을 내일로 이어가려면 무엇이 필요할까? GPT-4o와의 대화에서 시작해 수많은 실험과 실패를 거쳐 NEO라는 AI 운영체제를 만들기까지, 제가 직접 겪은 과정을 기록한 이야기입니다.

Build Your School's Software with AI* is a practical, plain-language guide for teachers and school staff. No programming experience is required. Everything is explained the first time it appears, every prompt is copy-paste ready, and every chapter ends with a checklist and a small exercise.

Your AI application works on your machine. What happens when it reaches production? Learn to deploy, monitor, secure, and scale AI systems using Docker, CI/CD, Kubernetes, cloud infrastructure, and practical production engineering. Move beyond prototypes and build AI applications you can operate with confidence.

Build agents that write code, fix tests, and commit changes while you sleep — safely, cheaply, and entirely on your own hardware (622 manuscript pages).

Que una IA tenga razón no significa que tenga permiso para actuar. Una guía práctica para diseñar sistemas de IA donde capacidad, evidencia, autorización y ejecución permanecen separadas antes de producir una consecuencia real.

AI can decide. That does not mean it is authorized to act. Governance Before Decision examines the missing control layer between intelligent systems and real-world consequences — where authority, scope, verification, and execution must meet.