AI Agents - The Definitive Guide

Regular price €76.99
Title
Quantity:
Will Deliver When Available
Will Deliver When Available
14 days return policy Shipping & Delivery
AI Agents LLM Agents Production AI Agent Deployment LLM Optimization GPU Optimization FlashAttention vLLM TGI LangGraph Agent Benchmarking LLM as a Judge Agent Evaluation AI Agent Security Guardrails MLOps for Agents Agent Orchestration Kubernetes AI Tool
Category=UYQE
eq_bestseller
eq_computing
eq_isMigrated=1
eq_nobargain
eq_non-fiction
forthcoming

Product details

  • ISBN 9798341666931
  • Dimensions: 178 x 232mm
  • Publication Date: 31 Oct 2026
  • Publisher: O'Reilly Media
  • Publication City/Country: US
  • Product Form: Paperback
Secure checkout Fast Shipping Easy returns

As AI agents move from research labs into production, engineering teams face mounting challenges—fragile tools, rising inference costs, erratic behavior, and unmet expectations. AI Agents: The Definitive Guide addresses what most books avoid: how to design, deploy, and maintain autonomous LLM agents that actually work in the real world. Written by Nicole Koenigstein, a leader in agentic AI with deep experience across finance, research, and engineering, this book offers the practical, system-level foundations needed to build robust, scalable, and secure agentic systems.

Whether you're tasked with making agent prototypes production-ready or building mission-critical automation from the ground up, this book guides you through every layer of the stack. It covers architectures, tool integration, performance optimization, safety strategies, and advanced evaluation, with a relentless focus on reliability and long-term value.

  • Design agent systems using modern frameworks like Reflexion, ReAct, and LangGraph
  • Optimize deployment with advanced model techniques and GPU-aware inference strategies
  • Securely integrate agents with tools using sandboxing, API contracts, and interface isolation
  • Build rigorous agent evaluation pipelines and address "agent-washing" failures
  • Scale fault-tolerant agents across orchestration platforms like Ray and Kubernetes
  • Apply debugging and monitoring practices to ensure safe agent reasoning and execution