Learn how to build GenAI applications using proven patterns and abstractions to work confidently across tools and frameworks
Free with your book: DRM-free PDF version + access to Packt's next-gen Reader*
Key Features
- Learn core abstractions and enterprise patterns behind modern GenAI systems
- Build secure, scalable agentic architectures using familiar engineering tools
- Generate multi-agent workflows and RabbitMQ-based GenAI configurations
- Use our Topologos prompt to easily build and test any agentic system
Book Description
Learn how to build GenAI applications using proven software engineering patterns instead of rapidly changing frameworks. This book helps engineers build secure, scalable agentic systems with familiar tools and practical, engineer-to-engineer architectural guidance.
You will connect GenAI concepts such as agentic workflows, embeddings, and vector databases to enterprise patterns, including components, adapters, and microarchitectures. Established GoF and enterprise design patterns help explain agentic behavior and system design, enabling you to reason about architecture rather than memorize tools. The book also shows you how to generate multi-agent and GenAI patterns as RabbitMQ configurations for scalable orchestration and communication.
Using language-agnostic examples and widely used messaging, orchestration, and data technologies, you will build production-ready systems that integrate with existing infrastructure without unnecessary complexity. You will also use our Topologos prompt to build, modify, and test deploy-ready multi-agent systems quickly while improving robustness and maintainability.
By the end of this book, you will be able to design reliable GenAI systems, make informed architectural decisions, and adapt confidently as tools and frameworks evolve.
*Email sign-up and proof of purchase required
What you will learn
- Build GenAI applications that are secure, reliable, and scalable
- Apply enterprise patterns to architect reliable GenAI systems
- Understand the core concepts behind LLMs needed for application development
- Use practical engineering approaches for agentic architecture design
- Generate RabbitMQ configurations for multi-agent GenAI systems
- Build and test robust multi-agent systems using our Topologos prompt
Who this book is for
This book is written for experienced software engineers, architects, and technical leads who want to add GenAI to their toolbox without abandoning the engineering principles that define good software.
It assumes familiarity with basic programming and software development, but does not require prior experience with AI, machine learning, RabbitMQ, or Airbyte. All necessary concepts and tools are introduced in contexts.
Table of Contents
- Introduction: Patterns, Abstractions, and the GenAI Landscape
- Embeddings: The Language of AI
- Building with GenAI: Parameters, Tuning, and Project Phases
- Building Your First RAG App
- Starting Your Data Migration Project
- Ingesting Data Using Airbyte and Pinecone
- Tips and Best Practices
- Pattern-Guided Coding: Using Patterns as the Design Vocabulary for GenAI Systems Built on RabbitMQ
- Implementing the ReAct Pattern Over RabbitMQ
- The Future and Limitations of LLMs
- Appendix A: Pattern Reference: GoF, Enterprise Integration, Reliability, and GenAI Microarchitecture Patterns
- Appendix B: Topologos User Manual
Tim O'Brien is CTO and Founder at two companies at the forefront of AI, blockchain, and decentralized technologies. His leadership has driven innovations across industries such as AI-powered document management, paperless cross-border trade, and digital identity. Tim has over 20 years of experience as a CTO, Architect, and Developer. His work with MiraScribe delivered advanced document automation solutions, while under his leadership, FuixLabs has excelled in no-code development, supply chain management, and fintech applications. Prior to his entrepreneurial ventures, Tim served as a Senior Engineer at Google, where he worked on language transpilation and distributed network systems. He also spent time at Hewlett-Packard, designing networking frameworks, and Cerner Corporation, where he created mission-critical, FDA-approved patient monitoring systems. Beyond his entrepreneurial success, Tim has consulted with firms like Augur, improving automated market-making and smart contract systems. He also founded Ferdon, a no-code platform for blockchain smart contracts, where he led the company from MVP development to securing venture funding.
