The AI Software Engineering Book

Soo Dong Kim, AI Software Engineering: Principles, Methods, and Practices, forthcoming, 2027.

This book presents a unified approach to engineering AI-based software systems throughout their lifecycle. Its scope includes traditional machine-learning-based systems as well as modern systems based on foundation models, generative AI, and agentic AI.

The book is not primarily a machine-learning textbook. Rather than focusing on the mathematics of algorithms or model training in isolation, it emphasizes the principles, methods, processes, artifacts, and best practices needed to engineer complete AI software systems. It covers the full lifecycle of AI systems, from process models and model engineering to software development, project management, and supporting activities.

Organization of the Book

The book is organized into six parts. Part I establishes the foundations of AI Software Engineering, and Parts II through VI address the engineering of AI systems from process, model, development, management, and supporting perspectives.

Part I — Foundations of AI Software Engineering
Introduces conventional software engineering foundations, AI systems and machine learning, and the principles, scope, challenges, and benefits of AI Software Engineering.

Part II — AI Software Process Models
Presents lifecycle frameworks for AI system development, including MLOps, hybrid development processes, and process tailoring.

Part III — AI Model Engineering Activities
Addresses the systematic engineering of data, models, prompts, and agent configurations, including preparation, training, validation, optimization, fine-tuning, versioning, monitoring, and evolution.

Part IV — AI Software Development Activities
Covers requirements specification, system analysis, data engineering, AI system architecture and design, model integration, implementation, testing, deployment, and maintenance.

Part V — AI Project Management Activities
Addresses project planning, effort and cost estimation, risk management, team management, and progress monitoring and control.

Part VI — AI Supporting Activities
Covers quality assurance, configuration management, metrics and measurement, process assessment and improvement, and documentation and communication for AI systems.

The book is intended for senior undergraduate and graduate students, software engineers, AI and ML engineers, data engineers, software and system architects, project managers, technical leaders, educators, and researchers who need a systematic engineering foundation for AI system development.

Supplementary materials will be provided through this website, including practical examples, engineering artifacts, templates, checklists, case-study materials, additional references, updates, and errata. These resources are intended to support university courses, self-study, and professional practice.