AI learning path
Learn AI step by step — from your first prompt to governed AI systems.
A simple map for the people who explain, sell and steer technology: three levels, twelve modules and one track per role. Every module links to the glossary terms it covers, so you can check any concept in a minute — and continue in the full course when you want lessons, exercises and feedback.
Four role tracks at a glance
Technical writers
From AI-assisted drafting to AI-ready content
Use AI to draft, review and translate — and make your documentation the trusted source that AI systems answer from.
Technical marketers
From prompts to visibility in AI answers
Produce on-brand content with AI, keep claims accurate and disclosed, and make your products visible in AI search.
Technical project managers
From pilots to governed AI delivery
Scope, budget and steer AI projects: choose models, set quality gates, manage data and vendor risk and keep people accountable.
Developers
Start at the expert level
Skim the beginner modules for shared vocabulary, then build with APIs, tools, agents and retrieval — and evaluate what you build.
Courses in the making
The four tracks are a preview — the courses are on their way.
The tracks above show where each role starts and which terms to learn first. The guided courses behind them — short lessons, role-specific exercises and quizzes, ending with a project from your own work — are being built right now. Send us a short email and we’ll let you know as soon as your track opens.
Three levels, twelve modules
Each module lists what you will be able to do afterwards and the terms behind it. “Core for” shows the roles that should not skip it.
Beginner
Understand and use
What AI is, who makes it and how to use it safely in daily work. No technical background needed.
B1
What AI is — and what it isn’t
Explain AI, machine learning, generative AI and LLMs in plain words, and why AI can be confidently wrong.
Core for: Technical writers · Technical marketers · Technical project managers
B2
Meet the AI landscape
Tell the major AI brands and their assistants apart and know which your company already uses.
Core for: Technical writers · Technical marketers · Technical project managers
B3
Your first good prompts
Write clear prompts with role, task, context, format and constraints, and share the ones that work.
Core for: Technical writers · Technical marketers · Technical project managers
B4
Use AI responsibly at work
Know what data may go into which tool, what must be reviewed and when AI use must be disclosed.
Core for: Technical writers · Technical marketers · Technical project managers · Developers
Intermediate
Apply and integrate
How models work one level deeper, prompting that scales across a team, and connecting AI to your own content.
I1
How models work, one level deeper
Reason about model types, reasoning modes, open weights and knowledge cutoffs when evaluating tools.
Core for: Technical project managers · Developers
I2
Prompting that scales
Build reusable system prompts, examples and structured outputs that give consistent results.
Core for: Technical writers · Technical marketers · Developers
I3
Ground AI in your content
Explain how retrieval-augmented generation works and what makes content ready for it.
Core for: Technical writers · Technical project managers · Developers
I4
Be found in AI answers
Make public product and documentation pages easy for AI answer engines to find, understand and cite.
Core for: Technical writers · Technical marketers
Expert
Design, build and govern
Choosing and combining models, building agents and tool integrations, and running AI with evaluation and governance.
E1
Choose and compare models
Run a structured model selection with your own test set, including open-weight and cloud options.
Core for: Technical project managers · Developers
E2
Build with APIs, tools and agents
Design an AI feature or agent with clear tools, permissions and human approval points.
Core for: Technical project managers · Developers
E3
Engineer the context
Decide what an AI system sees at each step — retrieval, metadata, memory — and measure the effect.
Core for: Technical writers · Developers
E4
Evaluate, secure and govern
Set quality gates with evals, defend against prompt injection and document AI for audits and regulators.
Core for: Technical writers · Technical marketers · Technical project managers · Developers
The AI landscape at a glance
The major AI brands and model families, each with what it offers and what it means for your role.
- OpenAI (ChatGPT, GPT models)OpenAI in brief: the company behind ChatGPT and the GPT models, its products for writers, marketers and developers, and what sets it apart.
- Anthropic (Claude)Anthropic in brief: the maker of Claude, its model tiers and apps, Claude Code, and its role in AI safety research and the MCP standard.
- Google DeepMind (Gemini)Google DeepMind and Gemini in brief: multimodal models across Search, Workspace and Android, open Gemma models, and developer access via Vertex AI.
- Meta (Llama)Meta and Llama in brief: open-weight language models under Meta’s community license, and the Meta AI assistant in Facebook, Instagram and WhatsApp.
- Microsoft (Copilot, Azure AI)Microsoft Copilot in brief: AI in Microsoft 365, Windows and GitHub, plus Azure AI Foundry for models from OpenAI and others.
- Mistral AIMistral AI in brief: the French developer of open-weight and commercial models and the Le Chat assistant — a European option.
- xAI (Grok)xAI and Grok in brief: the Grok models and assistant, their integration with X, and what to consider for brand and content use.
- DeepSeekDeepSeek in brief: the Chinese developer of open-weight reasoning models, why R1 mattered, and what to check before using its app or API.
- Alibaba Cloud (Qwen)Alibaba Cloud’s Qwen in brief: a broad family of open-weight language and multimodal models, from small to very large.
- Amazon (Bedrock, Nova)Amazon Bedrock and Nova in brief: managed access to many AI models on AWS, Amazon’s own Nova models and the Amazon Q assistants.
- NVIDIANVIDIA in brief: the GPUs and CUDA platform behind most AI training and inference, plus its own Nemotron models.
- CohereCohere in brief: the Canadian enterprise AI company whose Command, Embed and Rerank models target search, RAG and private deployment.
- PerplexityPerplexity in brief: the AI answer engine that cites its sources — and why it matters for content discoverability.
- Hugging FaceHugging Face in brief: the hub for open AI models, datasets and demos, and the Transformers library developers build on.
- Apple IntelligenceApple Intelligence in brief: on-device generative AI in iPhone, iPad and Mac, Private Cloud Compute and its privacy positioning.
- IBM (watsonx, Granite)IBM watsonx and Granite in brief: enterprise AI platform, open Granite models under Apache 2.0, and a strong focus on AI governance.
- Baidu (ERNIE)Baidu ERNIE in brief: China’s search leader and its ERNIE models and assistant — relevant for content and marketing in China.
- MidjourneyMidjourney in brief: the text-to-image generator popular for concept art and campaign visuals, and what to check on rights and disclosure.
- Adobe FireflyAdobe Firefly in brief: generative images and video inside Adobe apps, trained on licensed content, with Content Credentials for provenance.