AI Coding Agents

AI that doesn't just answer. AI that gets work done.

AI agents can plan, use tools, write and test code, search the web, query databases and check their own work. This practical guide shows you how to build them properly: step by step, with working Python code, clear diagrams and the safeguards real systems need.

Inside this book you will learn:

• The foundations: how large language models power agents, tool calling, context engineering and the ReAct loop
• Smarter agents: planning, reflection and self-correction, RAG, memory and the Model Context Protocol (MCP)
• Teams of agents: multi-agent design and orchestration, coding agents, browser agents, and voice and multimodal agents
• Production skills: evaluation, observability, security and guardrails, deployment, and cost and performance at scale

Five complete projects:
1. Deep Research Agent
2. Enterprise Knowledge Agent
3. SQL Data Analyst Agent
4. Autonomous Coding Agent
5. Multi-Agent Business Intelligence Team

Every project includes tested code, the design decisions behind it and the checks that make it reliable.

Who this book is for: developers, data engineers, students and technical professionals who want to move from simple chatbots to AI agents that are useful, reliable and safe.

About the author: Vivek Kumar Gupta holds a Master's degree in Artificial Intelligence and Machine Learning and has worked on many hands-on AI/ML projects.