Complementary books
Understand, master and humanize artificial intelligence. The books presented here explore AI from two complementary angles: educational and practical impact on the one hand, technical foundations and LLM engineering on the other.
The Mechanics of LLMs (technical level)
An engineering-focused book that dissects modern language models (Transformer, variants, agentic ecosystems) with an IT decision-maker's perspective.
- 15 progressive chapters: fundamentals, architecture, learning, agents and RAG
- 9 Python scripts to illustrate attention, RAG, agents and fine-tuning (LoRA)
- 153 pages oriented toward decisions: performance / cost / quality trade-offs
Audience: engineers, data scientists, advanced students, technical leaders.
One of the rare books offering a complete, technical and operational view of LLMs, far from simplifications and hype.
The Internal Mechanics of LLMs
Understand the mechanisms, build tools, check their results
Volume II by Mustapha Alouani: build tools around LLMs, measure their results and put conclusions to the test.
- Define the task and measurement criteria
- Build, inspect and test tools
- Distinguish observation from causal explanation
For readers who want to build and evaluate tools: engineers, developers, advanced students and technical leaders. Python and familiarity with tables help reproduce the experiments.
Paperback on Amazon
Kindle on Amazon
AI & Teaching (educational level)
A guide to understanding the impact of AI on education and to supporting pupils and students in their everyday use of AI tools.
- 10 practical chapters and a toolkit tested in educational contexts
- Focus on real use cases, risks and ethical reference points
- 235 pages accessible with no technical prerequisites
Audience: teachers, students, trainers, parents, educational decision-makers.
An accessible guide to support the educational digital transition, with concrete benchmarks and proven tools.
At a glance
| Aspect | The Mechanics of LLMs | The Internal Mechanics of LLMs | AI & Teaching |
|---|---|---|---|
| Main audience | Engineers, data scientists, CTO/CIO | Developers, engineers, advanced students, researchers | Teachers, students, parents |
| Objective | Understand LLM architecture and operation | Build tools, experiment and verify results | Understand and integrate AI in education |
| Level | Intermediate to advanced | Intermediate to advanced; Python helps reproduce experiments | Accessible, no technical prerequisites |
| Length (paperback) | 153 pages | 271 pages | 235 pages |
Page counts refer to the paperback editions presented; they may vary by language and format.
Which book for which profile?
- To understand architectures, attention, training and language-model applications: The Mechanics of LLMs — volume I, focused on foundations.
- To build tools, run experiments, inspect models and evaluate results: The Internal Mechanics of LLMs — volume II, focused on investigation.
- For an introduction to AI in education and guidance for everyday use: AI & Teaching.
- To connect technical understanding, experimental verification and educational support: the three books complement each other.
These books build on more than twenty years of experience in information systems and educational support, with the same requirement: putting technology at the service of people.