Agentic Architectural Patterns for Building Multi-Agent Systems

Transform GenAI experiments into production-ready intelligent agents with scalable AI systems, architectural patterns, frameworks, and responsible AI and governance best practices

Free with your book: DRM-free PDF version + access to Packt's next-gen Reader*

Key Features

• Build robust single and multi-agent GenAI systems for enterprise use
• Understand the GenAI and Agentic AI maturity model and enterprise adoption roadmap
• Use prompt engineering and optimization, various styles of RAG, and LLMOps to enhance AI capability and performance
• Purchase of the print or Kindle book includes a free PDF eBook

Book Description

Generative AI has moved beyond the hype, and enterprises now face the challenge of turning prototypes into scalable solutions. This book is your guide to building intelligent agents powered by LLMs.

Starting with a GenAI maturity model, you’ll learn how to assess your organization’s readiness and create a roadmap toward agentic AI adoption. You’ll master foundational topics such as model selection and LLM deployment, progressing to advanced methods such as RAG, fine-tuning, in-context learning, and LLMOps, especially in the context of agentic AI. You'll explore a rich library of agentic AI design patterns to address coordination, explainability, fault tolerance, and human-agent interaction. This book introduces a concrete, hierarchical multi-agent architecture where high-level orchestrator agents manage complex business workflows by delegating entire sub-processes to specialized agents. You’ll see how these agents collaborate and communicate using the Agent-to-Agent (A2A) protocol.

To ensure your systems are production-ready, we provide a practical framework for observability using life cycle callbacks, giving you the granular traceability needed for debugging, compliance, and cost management. Each pattern is backed by real-world scenarios and code examples using the open source Agent Development Kit (ADK).

*Email sign-up and proof of purchase required

What you will learn

• Apply design patterns to handle instruction drift, improve coordination, and build fault-tolerant AI systems
• Design systems with the three layers of the agentic stack: function calling, tool protocols (MCP), and A2A collaboration
• Develop responsible, ethical, and governable GenAI applications
• Use frameworks such as ADK, LangGraph, and CrewAI with code examples
• Master prompt engineering, LLMOps, and AgentOps best practices
• Build agentic systems using RAG, fine-tuning, and in-context learning

Who this book is for

This book is for AI developers, data scientists, and professionals eager to apply GenAI and agentic AI to solve business challenges. A basic grasp of data and software concepts is expected. The book offers a clear path for newcomers while providing advanced insights for individuals already experimenting with the technology. With real-world case studies, technical guides, and production-focused examples, the book supports a wide range of skill levels, from learning the foundations to building sophisticated, autonomous AI systems for enterprise use.

Table of Contents

1. GenAI in the Enterprise: Landscape, Maturity, and Agent Focus
2. Agent-Ready LLMs: Selection, Deployment, and Adaptation
3. The Spectrum of LLM Adaptation for Agents: RAG to Fine-tuning
4. Agentic AI Architecture: Components and Interactions
5. Multi-Agent Coordination Patterns
6. Explainability and Compliance Agentic Patterns
7. Robustness and Fault Tolerance Patterns
8. Human-Agent Interaction Patterns
9. Agent-Level Patterns

(N.B. Please use the Read Sample option to see further chapters)

Dr. Ali Arsanjani is a pre-eminent technical executive who bridges architectural rigor and large-scale organizational strategy with industrial-scale execution. Widely recognized as the "father of SOA", he has led transformational initiatives across multiple organizations. He currently serves as Director of Applied AI Engineering at Google Cloud, where he leads the GenAI Blackbelts, a center of excellence that bridges research, forward-deployed engineering, and enterprise implementation. In this role, he drives strategic co-engineering programs with Google's most critical customers and partners, accelerating enterprise adoption of generative AI and agentic AI. With executive roots as Head of Machine Learning at AWS and CTO for Analytics at IBM, Dr. Arsanjani has managed global teams of more than 6,000 practitioners. An IBM Master Inventor, his patent portfolio includes foundational contributions to service decomposition and context-aware routing. With executive roots as Head of ML at AWS and CTO for Analytics at IBM, Dr. Arsanjani has managed global teams of over 6,000 practitioners. An IBM Master Inventor, his patent portfolio includes foundational contributions to service decomposition and context-aware routing. Today, he pioneers generative AI and agentic AI, introducing best-practices and protocols (e.g., A2A) to ensure autonomous systems remain transparent, auditable, and grounded in ethical governance. A dedicated educator, he has impacted thousands of students as an Adjunct Professor at Maharishi International University, the University of California, San Diego, and San José State University. His scholarly work, cited more than 4,800 times, spans enterprise architecture, distributed systems, and neuro-machine interaction. Dr. Arsanjani remains a leading force in synthesizing deep technical mastery with the strategic vision required to navigate the Agentic Age.

Juan Pablo Bustos is a forward-thinking technology leader at the forefront of the generative AI revolution. With a distinguished background at industry giants including Google, Stripe, and Amazon Web Services, Juan specializes in operationalizing Artificial Intelligence for the enterprise. Currently at Google, he serves as a strategic partner to Fortune 50 corporations and global institutions, guiding them through the complex lifecycle of agentic AI adoption—from identifying high-impact use cases to deploying multi-agent systems at scale. Juan possesses the unique ability to zoom in and out of complex challenges, seamlessly translating high-level business strategy into rigorous technical architecture. He is passionate about empowering organizations to move beyond experimentation and deliver transformative value through cutting-edge technology.