Build an AI Agent (From Scratch)

Build an AI Agent (From Scratch): Agents that reason, plan, and act autonomously by Jungjun Hur and Younghee Song is a practical, step-by-step guide to AI agent development, LLM agents, Agentic AI, and autonomous AI systems. The book teaches Python developers and AI practitioners how to build a sophisticated AI agent from the ground up while understanding what happens inside modern agentic systems.

Instead of relying entirely on black-box AI agent frameworks, this book takes you through the underlying mechanics of building an AI agent from scratch. You will learn how LLMs can reason about complex tasks, select and use tools, retrieve information, maintain memory, plan actions, reflect on results, and work with other specialized agents.

Starting with the fundamentals of LLM-powered AI agents, you will learn how to distinguish traditional developer-defined workflows from LLM-directed agent loops. The book introduces important concepts such as agent loops, structured outputs, Pydantic, tool calling, context engineering, reasoning, planning and evaluation.

You will then build your agent's capabilities step by step using Python, including web search, local file exploration, vector-based RAG (Retrieval-Augmented Generation) and long-term memory. The book also explores Agentic RAG, allowing AI agents to retrieve and use relevant information while solving complex, multi-step problems.

A major focus is on context engineering, helping you understand how to structure information effectively and reduce problems such as context rot and the "Lost in the Middle" effect. You will also learn how AI agents can remember previous failures, maintain state and adapt their behaviour over time.

The book introduces the Model Context Protocol (MCP) for connecting AI agents with external tools and servers. You will learn how tool calling and external integrations can extend the capabilities of an LLM agent beyond simple text generation.

As your AI agent becomes more sophisticated, you will explore planning, procedural reflection and self-correction to help prevent hallucinations and endless agent loops. The book also introduces the Code-Act paradigm, allowing an agent to securely write and execute code in sandboxed environments.

The final stages explore multi-agent systems and agent orchestration. You will learn how specialized AI agents can collaborate through the A2A protocol, with tasks distributed among multiple agents to improve efficiency and solve complex problems.

The book also covers AI agent observability and evaluation, including LLM-as-a-judge approaches and OpenTelemetry-based tracing. This provides practical insight into monitoring, evaluating and improving AI agent performance.

By building the complete system from the ground up, Build an AI Agent (From Scratch) provides a practical foundation for anyone who wants to understand and develop modern AI agents, autonomous agents, LLM applications and multi-agent systems.

What You Will Learn

• Build an autonomous AI agent from scratch using Python
• Understand how LLM agents reason, act and make decisions
• Implement ReAct (Reasoning/Action) agent loops
• Use dynamic AI tool calling
• Apply structured outputs with Pydantic
• Understand and apply context engineering
• Build Agentic RAG and vector-based retrieval
• Implement AI agent memory and state persistence
• Add web search and external tools to AI agents
• Use MCP (Model Context Protocol) with AI agents
• Implement planning, reflection and self-correction
• Build sandboxed AI code execution systems
• Design and orchestrate multi-agent systems
• Explore the A2A protocol
• Implement AI agent observability and evaluation
• Use LLM-as-a-judge evaluation techniques
• Trace and improve AI agent performance with OpenTelemetry

What's inside

• End-to-end construction of a framework-free, autonomous AI agent in Python
• Advanced practical integrations including MCP, vector-based RAG, long-term memory, and sandboxed code execution
• Multi-agent orchestration and automated LLM-as-a-judge evaluation pipelines

Who This Book Is For

This book is aimed at Python developers, AI developers, machine learning practitioners and AI engineers who want to understand how AI agents work internally and learn how to build agentic systems without depending completely on abstract frameworks.

If you want to learn AI agent development, LLM agents, Agentic AI, autonomous AI systems, RAG, MCP, AI memory, tool calling or multi-agent architectures, this book provides a hands-on approach to building these technologies from first principles.

The publisher confirms that the book focuses on building a working AI agent step by step, including ReAct loops, MCP, Agentic RAG, memory, planning, specialized agents and multi-agent systems

Table of Contents

Part 1
1 What is an AI agent?
2 The brain of AI agents: LLMs
3 Enabling actions: Tool use
4 Implementing a basic ReAct agent
Part 2
5 Building knowledge bases with RAG
6 Adding memory to your agent
7 Planning and reflection for complex tasks
8 Empowering agents with code execution
9 Orchestrating multi-agent systems
10 Evaluating agents
A OpenAI API Key

Ebook License

End-User Warranty And License Agreement

1. Grant Of License

Manning Has Authorized The Download By You Of An Unrestricted Number Of Copies Of The Electronic Book (Ebook) In Any Of The Available Formats. Manning Grants You A Nonexclusive, Nontransferable License To Use The Ebook According To The Terms And Conditions Herein. This License Agreement Permits You To Install The Ebook On Any And All Your Devices For Your Personal Use Only.

2. Restrictions

You Shall Not: (1) Share, Resell, Rent, Assign, Timeshare, Distribute, Or Transfer All Or Part Of The Ebook Or Any Rights Granted Hereunder To Any Other Person; (2) Duplicate The Ebook, Except For A Single Backup Or Archival Copy; (3) Remove Any Proprietary Notices, Labels, Or Marks From The Ebook; (4) Transfer Or Sublicense Title To The Ebook To Any Other Party.

3. Intellectual Property Protection

The Ebook Is Owned By Manning And Is Protected By United States And International Copyright And Other Intellectual Property Laws. Manning Reserves All Rights In The Ebook Not Expressly Granted Herein. This License And Your Right To Use The Ebook Terminate Automatically If You Violate Any Part Of This Agreement. In The Event Of Termination, You Must Remove The Original And Any Copies Of The Ebook From All Your Devices.

4. Source Code Supplementary Material

Any Source Code Files Provided As A Supplement To The Book Are Freely Available To The Public For Download. Reuse Of The Code Is Permitted, In Whole Or In Part, Including The Creation Of Derivative Works, Provided That You Acknowledge That You Are Using It And Identify The Source: Title, Publisher And Year.

5. Limited Warranty

Manning Warrants That The Ebook Files, A Copy Of Which You Are Authorized To Download, Are Free From Defects In The Operational Sense That They Can Be Read By A Pdf Reader Or Epub Reader, Or Other. Except For This Express Limited Warranty, Manning Makes And You Receive No Warranties, Express, Implied, Statutory Or In Any Communication With You, And Manning Specifically Disclaims Any Other Warranty Including The Implied Warranty Of Merchantability Or Fitness Or A Particular Purpose. Manning Does Not Warrant That The Operation Of The Ebook Will Be Uninterrupted Or Error Free. If The Ebook Was Purchased In The United States, The Above Exclusions May Not Apply To You As Some States Do Not Allow The Exclusion Of Implied Warranties. In Addition To The Above Warranty Rights, You May Also Have Other Rights That Vary From State To State.

6. Limitation Of Liability

In No Event Will Manning Be Liable For Any Damages, Whether Arising For Tort Or Contract, Including Loss Of Data, Lost Profits, Or Other Special, Incidental, Consequential, Or Indirect Damages Arising Out Of The Use Or Inability To Use The Ebook.

7. General

This Agreement Constitutes The Entire Agreement Between You And Manning And Supersedes Any Prior Agreement Concerning The Ebook. This Agreement Is Governed By The Laws Of The State Of New Yor