Building Agentic and Autonomous AI Agents: Design Intelligent LLM Agents with Tools, Memory, Reasoning, Planning, and Multi-Agent Workflows - Softcover

Carlson, Abe

9798192280355: Building Agentic and Autonomous AI Agents: Design Intelligent LLM Agents with Tools, Memory, Reasoning, Planning, and Multi-Agent Workflows

Softcover

ISBN 13: 9798192280355

Publisher: Independently published, 2026

View all copies of this ISBN edition

0 Used

2 New

From US$ 25.00

AI is moving beyond systems that simply generate text. The next generation of intelligent applications can interpret goals, reason about tasks, use tools, maintain context, make decisions, and execute multi-step workflows.

But building reliable AI agents requires more than connecting an LLM to a prompt.

Building Agentic and Autonomous AI Agents provides a practical engineering guide to designing, developing, evaluating, securing, and deploying modern agentic AI systems. It takes you from the foundations of AI agents to autonomous workflows, tool use, memory, reasoning, planning, orchestration, and multi-agent architectures.

You will learn how to:

Understand the architecture and core principles behind agentic AI
Design LLM-powered agents that can reason and make decisions
Build agents that use tools, APIs, databases, files, and external services
Implement short-term, long-term, and persistent agent memory
Design task decomposition, planning, reflection, and adaptive reasoning loops
Build sequential, conditional, iterative, event-driven, and human-in-the-loop workflows
Move from predefined workflows toward controlled autonomous execution
Design multi-agent systems with specialized roles, delegation, coordination, and supervision
Connect retrieval-augmented generation to AI agents
Test agent decisions, tool usage, reliability, and task completion
Protect agents against prompt injection, unauthorized actions, and unsafe tool use
Monitor, evaluate, optimize, deploy, and scale production-ready agentic applications
Understand advanced architectures such as ReAct-style, planning-based, reflection-based, memory-augmented, and multi-agent systems

Rather than treating AI agents as mysterious autonomous systems, this book approaches them as engineered software systems with models, state, tools, memory, control loops, policies, and measurable outcomes.

Whether you are a developer exploring LLM agents, an AI engineer building production systems, or a technical professional seeking to understand autonomous AI architectures, this book provides a structured path from LLM applications to intelligent, tool-using, reasoning, and autonomous systems.

This is Book 1 of The Agentic AI and GraphRAG Engineering Series.

It establishes the foundations you will need before moving into the knowledge-aware architectures explored in Book 2, Building Intelligent AI with GraphRAG.

"synopsis" may belong to another edition of this title.