The 28-day AI curriculum
28 lessons of about twelve minutes each, grouped into four weeks. Every day comes with a narrated slide deck, a comic storyboard, copy-paste prompt templates and a three-question quiz. After Day 28: a 50-question, 60-minute exam — 35 correct earns the certificate.
Week 1: Foundations
Days 1–7
How models actually work, the anatomy of a strong prompt, context windows, sampling, few-shot examples and structured reasoning.
- Day 1 · 12 min
Demystifying LLMs
What really happens when you press enter
- Day 2 · 14 min
The Anatomy of a Perfect Prompt
Persona, Context, Task, Constraints
- Day 3 · 11 min
Context Windows & Working Memory
Why the model 'forgets', and how to stop it
- Day 4 · 10 min
Temperature, Sampling & Determinism
Choosing between creative and consistent answers
- Day 5 · 12 min
Few-Shot Prompting by Example
Show the pattern instead of describing it
- Day 6 · 13 min
Chain-of-Thought & Structured Reasoning
Give the model room to work
- Day 7 · 15 min
Week 1 Consolidation: Your Prompt Library
Save the prompts that work so you can use them again
Week 2: Applied Productivity
Days 8–14
Writing, inbox triage, research synthesis, meeting extraction, spreadsheets and decks — the tasks that fill an actual working week.
- Day 8 · 13 min
AI for High-Velocity Writing
Drafts are cheap; choosing is the real work
- Day 9 · 11 min
Inbox Zero with AI Triage
Decide once, then reply fast
- Day 10 · 14 min
Research & Synthesis at Speed
Understand fast, but keep your own judgement
- Day 11 · 10 min
Turning Meetings into Momentum
From meeting notes to clear next steps
- Day 12 · 13 min
Spreadsheets, Data & Analysis
Let the model write the formula, not compute the answer
- Day 13 · 12 min
Slides and Narrative Structure
Get the story right before the design
- Day 14 · 15 min
Week 2 Consolidation: The Productivity Audit
Find where AI really saves you time
Week 3: Workflows & Automation
Days 15–21
Prompt chaining, JSON schemas, retrieval-augmented generation, tool use, agents with guardrails and your first real automation.
- Day 15 · 13 min
Prompt Chaining & Task Decomposition
Many small steps beat one giant prompt
- Day 16 · 12 min
Structured Output: JSON, Tables & Schemas
How AI answers can feed into other software
- Day 17 · 14 min
Retrieval-Augmented Generation (RAG)
Making the AI answer from your own documents
- Day 18 · 13 min
Tool Use & Function Calling
Letting the AI do things, not just talk
- Day 19 · 14 min
Agents: Autonomy with Guardrails
When letting the AI work alone is worth the risk
- Day 20 · 12 min
Automating Recurring Workflows
From a prompt you run by hand to a system that runs itself
- Day 21 · 16 min
Week 3 Consolidation: Build Your First Workflow
Launch one thing that runs without you
Week 4: Scale, Safety & Strategy
Days 22–28
Hallucination discipline, bias, data boundaries and prompt injection, evaluation, model economics and rolling AI out to a team.
- Day 22 · 13 min
Hallucinations & Verification Discipline
Why sounding sure and being right are not the same
- Day 23 · 12 min
Bias, Fairness & Responsible Use
How AI can repeat unfairness that already exists
- Day 24 · 13 min
Security, Privacy & Data Boundaries
What you paste into AI matters
- Day 25 · 14 min
Evaluating AI Output Like an Engineer
Replace 'that looks good' with a real test
- Day 26 · 12 min
Cost, Latency & Model Selection
The best model is the cheapest one that does the job
- Day 27 · 13 min
Rolling AI Out Across a Team
Getting a team to use AI is about people, not tools
- Day 28 · 18 min
Capstone: Your AI Operating System
Twenty-eight days, put together into one way of working