Artificial Intelligence

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

Skip the black-box frameworks. Build a production-grade AI coding agent from scratch in pure Python - cloud or local, tested with pytest, all in a single file.

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

You don’t need massive compute or big-tech resources to build real AI. This hands-on guide shows you how to build and fine-tune your own small language model—from scratch—using accessible tools like Google Colab. Learn transformers step-by-step, train and align models on your own data, and deploy practical AI systems that run on consumer GPUs. If you’ve ever wondered “Can I build my own model?”—this book proves the answer is yes.

Language models guess. Knowledge graphs know. Rusty Graph shows you how to build a local AI agent whose memory is a real knowledge graph — typed, validated, reasoned over, and queryable with SPARQL — all inside a single static Rust binary. No JVM. No Docker sidecar. No Python runtime bolted to the side. You will build Ares a research-assistant agent that observes papers, forms beliefs, makes promises to other agents, and tracks the provenance of every fact it holds. Chapter by chapter, Ares grows from an empty Cargo workspace into a full pipeline: > load → reason → validate → query → answer All of it in under a thousand lines of idiomatic Rust, using three crates that actually work today: `oxigraph`, `reasonable`, and `rudof_lib`. `grapfeo` You will learn how to- Model a domain as RDF triples and load them into an embedded store. - Write RDFS and OWL 2 RL axioms that infer trust, identity, and inverse relationships — automatically. - Guard your graph with SHACL shapes that reject bad data at the boundary, not in production. - Query everything with SPARQL, from simple lookups to federated queries across named graphs. - Wire the graph into a hybrid RAG pipeline so your LLM answers are grounded in facts, not vibes. Who it is forRust developers building agents, assistants, or any system where the answer "the model said so" is not good enough. You should be comfortable with Cargo and traits. You do **not** need any prior semantic-web background — every concept is introduced through Ares before any formal definition appears. Why Rust, why nowLocal agents are the next deployment target: a user's laptop, a Raspberry Pi, a WASM sandbox. Python cannot go there comfortably. Rust can. This book is the missing manual for the Rust side of the semantic web — the one that tells you exactly which crates work, where the ecosystem is thin, and how to ship anyway. Stop hoping your model tells the truth. Give it a graph that does.

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

Run Jev-style decision models on your own hardware with Ollaya. Learn how these fast, structured models differ from LLMs, then build real applications with typed outputs, Python, JavaScript and MCP. From first setup to production, this practical guide shows you how to make AI faster, cheaper and more private.

Build a complete LLM inference engine in C++ — from a blankproject to a working Transformer that loads a real model andgenerates text. Forged one challenge at a time, with teststhat prove every piece works before you move on.

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.

A leader at a company opens Slack one morning. The brief their virtual employee wrote looks normal: same headers, same prose voice, same cadence as every other brief that week. The last line ends mid-sentence. The API returned `stop_reason: "max_tokens"`. The system shipped it anyway. No exception. No log line. No retry. That bug doesn't look like the normal kind. This book is about why, and what to build instead.

Securing AI requires a system-level approach across models, data, identities, tools, infrastructure, and execution paths. This book shows how to apply Product Security and Secure SDLC principles to AI-enabled products, from RAG and agents to MCP, evaluations, supply chain risk, runtime controls, and governance, while building the fundamentals needed for a practical AI Product Security operating model.

Reverse engineering gets a powerful upgrade with Claude Code. Learn how to investigate, understand and reconstruct software you’re authorized to analyze, while keeping evidence, accuracy and reproducibility at the center. From legacy systems to modern codebases, this book turns AI into a practical partner for serious software analysis.

An LLM is not an AI system.Systems Thinking for Agentic AI shows software engineers and architects how to design reliable AI applications with prompts, RAG, tools, memory, orchestration, guardrails, evaluation, observability, and runtime control.Move beyond chatbot demos and learn how to build production-ready agentic AI systems you can reason about, measure, debug, operate, and improve

Learn how to build, run, and optimize llama.cpp from the ground up. This book covers everything from compiling the code and working with GGUF models to deploying fast, production-ready local LLM inference.

The most important impact of AI on software development isn't that it writes code faster — it's that it changes what you can delegate. Agentic Programming is a practical roadmap for climbing the AI Fluency Ladder: from prompting to agentic workflows, verified execution, and ultimately autonomous software development.