Artificial Intelligence

A clear, illustrated guide to large language models, covering key concepts and practical applications. Ideal for projects, interviews, or personal learning.

Bestselling book on building LLMs. A practical, project-driven manual for engineers who want to understand how modern language models are built — and where they fail — by writing every layer themselves. From a scalar autograd engine to RLHF to fused specialists, in 36 hands-on projects with deliberate sabotage experiments. Build it. Break it. Measure it.

Your use cases have a new reader: an AI agent that turns them directly into code and tests. Implementation is now cheap but expressing intent precisely is the bottleneck. This guide teaches you to write use cases that work for stakeholders, engineers, and the AI agents that implement from them.

Master language models through mathematics, illustrations, and code―and build your own from scratch!

Everything you really need to know in Machine Learning in a hundred pages.

Revised for PyTorch 2.x! In 2019, I published a PyTorch tutorial on Towards Data Science and I was amazed by the reaction from the readers! Their feedback motivated me to write this book to help beginners start their journey into Deep Learning and PyTorch. I hope you enjoy reading this book as much as I enjoy writing it.

It's never been easier to build an AI agent — and never been harder to make one that actually works. This book takes you from language model foundations to production-ready multi-agent systems with the depth to predict failure before it happens, engineer graceful degradation over catastrophic failure, and take absolute architectural ownership. Get the paperback from amazon.

Your C# skills are worth more today than they were a year ago — if you know how to put a language model in the loop. This book shows you how, with the Microsoft Agent Framework: real tools, RAG, multi-agent orchestration, plus the hosting, observability, and safety that separate a demo from a system you ship. Nineteen chapters. 120 runnable projects. No Python detours. Just C# and .NET 10.

A model that crashes is a good day. The dangerous failures return answers — plausible, fluent, and wrong. 100 failures. 63 diagnostic instruments. One rule: inspect what actually happened.

A practical guide to fine-tuning Large Language Models (LLMs), offering both a high-level overview and detailed instructions on how to train these models for specific tasks.Get the paperback version here. Get the Kindle version here.

Modern Deep Learning models can be extremely large, often exceeding the memory capacity of a single GPU or CPU. In these cases, training must be distributed across multiple processors. This introduces the need for high-speed communication between GPUs—both within a single server and across multiple servers. Intra-node GPU communication typically relies on high-speed interconnects like NVLink, with Direct Memory Access operations enabling efficient data transfers between GPUs. Inter-node communication, however, depends on the backend network, either  InfiniBand or Ethernet-based. Synchronization of model parameters across GPUs places strict requirements on the network: high throughput, ultra-low latency, and zero packet loss. Achieving this in an Ethernet fabric is challenging but possible.   This is where datacenter networking meets Deep Learning. Understanding how GPUs communicate and what the network must deliver is essential for designing effective AI data center infrastructures.

AI governance is moving from principle to practice. This hands-on guide shows you how to build, implement and certify an ISO/IEC 42001 AI management system with confidence. From clause-by-clause guidance to practical templates, integration strategies and real-world scenarios, it turns a complex standard into a clear path from foundation to certification.

How to represent knowledge for LLMs and build memory for agents, I discovered Mark's work on semantic spacetimes. It's more of a theoretical framework from someone who came from physics. But actually, jumping to knowledge representation and reasoning, and trying to answer the question of how to build dynamic and complex systems—semantic spacetimes and promise theories are crucial for the future of agentic systems, in my belief.The semantic spacetime approach gives us answers on how to organize better memory and how to have better knowledge representation that could be understood quite well by LLMs. Vector embeddings actually create a lot of challenges—some spaces and some relations in vector embeddings simply don't exist. We all have this problem where "love my wife " and "hate my wife" while actually quite distant in practice, and also time and dynamics matter

Unlock the power of AI in your applications with this groundbreaking book on AI-driven application architecture. Discover practical patterns and principles for building intelligent, adaptive, and user-centric software systems that harness the potential of large language models and AI components.

Move beyond individual prompts and engineer the full information environment an AI model receives. Learn to design, test, measure, and govern context for reliable AI workflows.