Skip to content
ISBN: B0HHWDFNCM
ISBN13: 9798172009013

AI Agents Engineering Handbook: Build, Deploy, and Scale Autonomous AI Agents with LLMs, RAG, Tools, and Workflows

ISBN: B0HHWDFNCM
ISBN13: 9798172009013
Formats

Conditions

Formats

Conditions

See all 1 editions

Book Overview

AI agents are more than chatbots with better prompts. They are software systems that can interpret goals, use tools, retrieve information, maintain state, make decisions, and carry out multi-step tasks. Building them well requires more than connecting an LLM to an API.

AI Agents Engineering Handbook provides a practical introduction to designing and building AI-agent systems, starting with the fundamentals and progressing toward evaluation, security, deployment, and monitoring.

You will learn how to:

  • Understand what AI agents are and how they differ from chatbots, traditional software, and fixed workflows.
  • Build an AI agent and understand the agent loop of goals, decisions, actions, observations, and completion.
  • Give agents controlled access to tools, APIs, databases, search systems, and application functions.
  • Design tool schemas, validate inputs, manage permissions, handle failures, and control tool execution.
  • Work with context, state, conversation history, and long-term memory.
  • Connect agents to external knowledge using Retrieval-Augmented Generation (RAG), embeddings, chunking, vector stores, retrieval, and source attribution.
  • Decide when an agent is appropriate and when a predictable workflow or ordinary software is the better solution.
  • Design planning and workflow systems with clarification, human approval, stopping conditions, and controlled execution.
  • Build practical agent architectures for customer support, research, data analysis, productivity, and document assistance.
  • Test agents beyond their final answers by evaluating tool selection, tool arguments, retrieval quality, safety, efficiency, reliability, and failure recovery.
  • Protect agents using least privilege, authorization, input validation, data minimization, prompt-injection defenses, human approval, rate limits, timeouts, and audit logging.
  • Move an agent from prototype toward production through deployment architecture, monitoring, logging, latency and cost management, retries, health checks, versioning, and rollback planning.
  • Bring the concepts together in a complete capstone AI-agent system with knowledge retrieval, state, tools, validation, human escalation, security checks, testing, and monitoring.

The book takes an engineering-first approach. Instead of treating autonomy as the goal, it focuses on designing agents around a real problem and giving them only the capabilities they actually need.

You will also learn why an agent that works during a demonstration is not automatically reliable in production, and why testing, security, observability, controlled permissions, and failure handling must be part of the design.

Whether you are beginning to explore agentic AI, entering AI engineering, building AI-powered applications, or looking to understand how modern AI-agent systems are structured, this book gives you a practical foundation for moving from "I understand what an AI agent is" to "I understand how to design and build one."

Format:
Paperback
Language:
English
ISBN:
B0HHWDFNCM
ISBN13:
9798172009013
Release Date:
September 2026
Publisher:
Independently Published
Length:
322 Pages
Weight:
1.65 lbs.
Dimensions:
0.7″ × 8.5″ × 11.0″

More by TERESA S. CURRY

Customer Reviews

0 customer ratings | 0 reviews
Write a Review

There are currently no reviews. Be the first to review this work.