Book 01 · Model foundations
Inside a GPT-Style Language Model
Tokens, attention, training and generation, assembled from first principles.
Read the book →AI Systems Library
A growing library of rigorous, visual field books for people who build, govern and operate AI systems.
18 accepted editions · 188 static Mermaid chapter maps
Book 02 · Reasoning systems
Engineering Reasoning Models
Train, evaluate and constrain systems that solve multi-step problems.
Read the book →Book 03 · Foundations
Applied Language Models
A practical route from prompts and embeddings to useful language systems.
Read the book →Book 04 · Security and privacy
Secure Language Model Systems
Threat models, privacy boundaries and controls for language-model applications.
Read the book →Book 05 · Agents and control
Controlled Agents
Design agent loops with visible state, bounded tools and human authority.
Read the book →Book 06 · Model foundations
Language Models, Piece by Piece
A code-first journey through the machinery of a modern language model.
Read the book →Book 07 · Small models
Small Models, Hard Evidence
Choose and adapt compact language models with disciplined measurement.
Read the book →Book 08 · Reasoning systems
Reasoning Under Test
Build reasoning systems whose improvement can survive honest evaluation.
Read the book →Book 09 · Serving
Serving Under Load
Turn trained models into reliable services under real latency and capacity pressure.
Read the book →Book 10 · LLMOps
The System After the Demo
Operate language models through evaluation, release, monitoring and recovery.
Read the book →Book 11 · RAG and evidence
The Evidence Is the Product
Design retrieval systems whose answers can carry their own proof.
Read the book →Book 12 · Model foundations
The Machine Behind the Sentence
Understand how language models turn text into probability, structure and output.
Read the book →Book 13 · Evaluation
The Model Is Not the Test
Design evaluations that measure the system, the route and the consequences.
Read the book →Book 14 · LangGraph
The Graph Is a Contract
State, routes, vetoes and receipts in controlled AI workflows.
Read the book →Book 15 · Agents and protocols
The Agent Has No Authority
Compose, persist and connect AI systems without surrendering control.
Read the book →Book 16 · Machine learning foundations
A Score Is Not a Decision
Machine learning from objective and evidence to monitored human judgement.
Read the book →Book 17 · ML system design
The Model Must Survive Contact
Machine-learning system design from decision contract to owned operation.
Read the book →Book 18 · Language model engineering
The Language Model Workbench
Inspect, adapt and evaluate language models from tokens to tuned systems.
Read the book →Visual publication standard
Authored figures and Mermaid maps, together
Each accepted edition retains its authored information designs and receives responsive, accessible Mermaid maps across its substantive sections. Completed editions are published here after the collection build and site checks pass.