Facing the challenge of orchestrating AI agents that think, act, and adapt across multiple services? "Designing Agentic AI Systems" delivers a complete blueprint for building autonomous workflows with real-world reliability.
At its core, this book reveals how to transform standalone language models into coordinated, goal-driven agents that call external APIs, manage state, and recover from errors. You'll discover proven design patterns (Chapter 2), a standardized Model Context Protocol (MCP) for seamless agent-tool communication (Chapter 3), and hands-on tutorials for LangGraph and AutoGen-two frameworks that turn ideas into production-ready pipelines (Chapters 4-6).
What You'll Master:
Agent-Driven Workflows: Learn task decomposition, dynamic replanning, and error-handling strategies that keep your agents running smoothly, even under failure conditions.
Interoperable Protocols: See how MCP message schemas enable transparent, auditable exchanges between agents and external services, ensuring every API call is logged, validated, and retried if necessary.
Graph-Based Orchestration with LangGraph: Build, test, and deploy stateful workflows (Chapter 4). Define nodes and edges to link LLM reasoning, tool invocation, and memory layers into a cohesive pipeline.
Multi-Agent Coordination via AutoGen: Define roles, create task-handoff scenarios, and share memory across agents (Chapter 5). Implement KV and vector memory layers for persistent context.
Performance and Scalability: Explore horizontal versus vertical scaling strategies, GPU batching, and caching techniques to optimize resource usage (Chapter 11).
Security, Compliance, and Ethics: Protect agent-tool communications with authentication and role-based authorization, safeguard sensitive data, and implement bias-detection checks (Chapter 10).
Whether you're an AI engineer, developer, or technical lead, this book equips you with the practical skills and code examples needed to architect agentic AI systems that are robust, scalable, and maintainable.
Take control of your AI workflows today, order "Designing Agentic AI Systems" and build autonomous agents your users can trust.
外文書商品之書封，為出版社提供之樣本。實際出貨商品，以出版社所提供之現有版本為主。部份書籍，因出版社供應狀況特殊，匯率將依實際狀況做調整。
無庫存之商品，在您完成訂單程序之後，將以空運的方式為你下單調貨。為了縮短等待的時間，建議您將外文書與其他商品分開下單，以獲得最快的取貨速度，平均調貨時間為1~2個月。
為了保護您的權益，「三民網路書店」提供會員七日商品鑑賞期(收到商品為起始日)。
若要辦理退貨，請在商品鑑賞期內寄回，且商品必須是全新狀態與完整包裝(商品、附件、發票、隨貨贈品等)否則恕不接受退貨。
