Learn practical AI skills for your work
With practical exercises
Detailed articles, 15 guides and resources for working with AI assistants. Learn through quizzes, diagrams and exercises, no account needed.
Free articles, exercises and resources.
One goal, one first exercise
[ WHY IT MATTERS ]
Guides to progress
From your first prompt to agents and workflows.
Required accounts
Learning content is available without signing up.
Content languages
Articles and guides in French and English.
[ LEARNING PATHS ]
15 free paths to practical skills
Choose a goal, complete an exercise and verify your result.
Foundations
Prompt Engineering Basics
Discover prompting fundamentals and build your first effective prompts with the R.C.T.F framework.
How LLMs Work
Understand tokens, attention, and generation mechanics. Build clear, reusable prompts.
Structured AI Outputs
Get reliable JSON, tables, and structured data from AI with schema-driven prompting.
Techniques
Chain-of-Thought Reasoning
Trigger complex reasoning with CoT, Self-Consistency, and Tree-of-Thought techniques.
Prompt Orchestration
Build multi-step workflows with prompt chaining, routing, and map-reduce patterns.
RAG Retrieval-Augmented Generation
Ground AI in reliable sources. Build document retrieval pipelines and context windows.
Application
[ LATEST ARTICLES ]
From the Blog
Prepare an association grant application with Claude: evidence, project and budget
Organize application requirements, draft a project description and reconcile a grant budget. French association example, practical lab and checklist.
Read →ClaudeSummarize a document folder with Claude: sources, conflicts and a useful report
Turn a folder of documents into a verifiable report. Track versions, cite evidence, expose contradictions and use a practical review checklist.
Claude, ChatGPT or Gemini for office work: compare them on your actual tasks
Use a reproducible evaluation for documents, spreadsheets and presentations: evidence, calculations, formats, correction time, cost and constraints.
Review learner work with AI: build a rubric and keep the final decision human
Prepare formative feedback with a transparent rubric, evidence from the work and calibration examples. Keep grading decisions with the educator.
Create your first Claude Skill: a complete meeting-notes example
Build a reusable meeting-notes Skill with a valid SKILL.md, explicit boundaries, fictional test cases, a practical lab and a downloadable template.
TypeSafe Jev: decision models, real benchmarks and limitations
Understand Jev typed decisions, independent Decision Index results, calibration limitations and a practical routing evaluation worksheet.
AI models in October 2026: OpenAI, Gemini, Claude, Meta, Grok, Mistral and DeepSeek
A dated guide to current AI model families, documented access, practical evaluation, benchmark limitations and cost per accepted result.
AI video models in 2026: compare Veo, Gemini Omni, Runway, Kling and LTX
Evaluate video models for consistency, motion, sound and cost per accepted shot. Create educational content without filming yourself.
[ NEWSLETTER ]
Stay Ahead in AI
Get our best articles, tutorials and guides delivered straight to your inbox. Free, no spam.
One practical AI workflow in your inbox
Methods, resources and new exercises. Free, unsubscribe at any time.
Ready to Level Up Your AI Skills?
Start with our beginner guide or dive into 150+ articles covering every aspect of AI engineering. No account, no paywall just start learning.
Frequently Asked Questions
Is everything really free? No hidden fees?+
Articles, 15 guides, quizzes, exercises and public resources are free. Personalized coaching is a separate service with an agreed quote.
Can I learn AI without coding?+
Absolutely. No programming skills are required to start. Our content is designed for all backgrounds — marketers, HR professionals, designers, students, or anyone curious about AI. The beginner guides (Modules 0-2) require zero code. Advanced guides like RAG and AI Agents introduce technical concepts progressively with hands-on exercises.
Is prompt engineering still relevant in 2026?+
Yes, but it’s evolving. Prompt engineering remains essential, but the field has expanded into context engineering — selecting, compressing, and orchestrating the right information for LLMs. Our curriculum covers both: from foundational prompt techniques (zero-shot, few-shot, chain-of-thought) to advanced context engineering (RAG, MCP protocol, caching strategies).
How do I learn prompt engineering step by step?+
Start with Guide 0 (Prompt Engineering Basics) to learn the R.C.T.F framework and core techniques. Then progress through LLM internals (Guide 1), structured outputs (Guide 2), and chain-of-thought reasoning (Guide 3). Each guide builds on the previous one with interactive exercises. The full path from beginner to expert takes about 40 hours at your own pace.
What topics does the curriculum cover?+
The 15 guides cover AI foundations, prompting, structured outputs, RAG, agents, safety and five complete courses.
What is context engineering vs prompt engineering?+
Prompt engineering focuses on crafting the right instruction for an LLM. Context engineering goes further — it’s about selecting, compressing, and managing ALL the information surrounding the prompt: retrieved documents (RAG), tool outputs, conversation history, and system instructions. Think of prompt engineering as writing a good question; context engineering is about curating everything the AI needs to give a great answer. We cover both in depth.
Is the content available in French?+
Articles and guides are available in French and English. Some interface pages are offered in other languages and point to available content.