Building AI Applications bundle

Designing AI Systems

AI applications need much more than a connection to a model. To work well in the real world, they need memory, access to company knowledge, safe ways to take action, clear rules, and reliable operations behind the scenes. Rather than rebuilding those pieces for every project, organizations can create a shared AI platform that supports all their AI applications.

Designing AI Systems shows you how to build that platform. Starting from scratch, you’ll create the core services that real AI applications depend on: a way to connect to different AI providers, a way to remember earlier interactions, a way to bring in your organization’s own data, and a way to let AI systems use tools and act safely within company rules. By the end, you’ll understand not just how to call an AI model but how to build the full system around it.

This book focuses on the part of AI development that usually gets overlooked: everything required to make an AI system reliable, manageable, and ready for production. With complete Python code in every chapter, you’ll build a working reference implementation you can adapt to your own needs. You’ll also learn the practical operational skills that matter in production, including managing costs, routing requests, enforcing safety controls, and evaluating system quality.

The result is a solid foundation for AI applications that can use company knowledge, carry out multi-step work, connect to outside tools, and remain trustworthy and governable as they scale.

Grokking AI Applications

“Clear, practical, and focused on getting you from basics to real-world applications fast.”
—Michele Pacifico, Nestlé

Grokking AI Applications: An illustrated guide for programmers and other curious people shows you how to create AI applications like chatbots, agents, and systems customized to use your own data—even if you’ve never built one before! Supporting his explanations with graphics and using concrete examples and hands-on implementations, author Andrea De Mauro gets you thinking about system design and mental models rather than boilerplate code and syntax debugging. You’ll quickly focus on the right level: how to manage prompts, data, and application components to get the results you want.

You’ll start with a bare-bones chatbot, as you learn how to create good prompts and improve accuracy using Retrieval Augmented Generation (RAG). To help your applications live in the real world, the book discusses how to connect them to document collections, structured databases, and CRM tools. You’ll build on this foundation, creating your own agents and equipping them with tools for data analysis and complex reasoning. Reviewer Julien Pohie of Thoughtworks commented, “The emphasis on multi-agent orchestration aligns perfectly with modern engineering best practices.”

Step-by-step guidance, rich diagrams, and workflow summaries help you stay oriented, even through advanced material. To emphasize conceptual and design thinking, you’ll start each project using the low-code development tool Langflow. For readers who want to go deeper, every example also includes fully-annotated Python implementations. You’ll see the code, libraries, APIs, and data behind each application, along with practical advice on how to deploy and monitor your applications.

Compact by design, this fast-paced book includes eight fully illustrated chapters that build intuition through engaging, memorable, and relatable real-world scenarios. As you read, you’ll experience a shift from AI as the familiar conversational chatbot to AI as an essential part of your application and, further, to the design of multi-agentic teams with specialized roles and supervised coordination.

And you’ll enjoy every page of the journey!

LLMs in Production

LLMs in Production teaches you how to develop an LLMOps plan that can take an AI app smoothly from design to delivery. You’ll learn techniques for preparing an LLM dataset, cost-efficient training hacks like LORA and RLHF, and industry benchmarks for model evaluation. Along the way, you’ll put your new skills to use in three exciting example projects: creating and training a custom LLM, building a VSCode AI coding extension, and deploying a small model to a Raspberry Pi.