01
Foundation
Think Like
an AI User
Understand what AI is, what it can and can't do, and how to get useful outputs
Core Topics
- What is AI? History: Narrow AI → Machine Learning → Deep Learning → Generative AI → LLMs
- Key vocabulary: token, model, parameter, context window, hallucination, inference
- How LLMs work — simply: next-token prediction explained in plain English
- The 3 non-negotiables: AI amplifies thinking, all outputs need verification, context quality determines quality
Graduation Project
AI Tool Comparison
Test 3 different AI tools. Ask each the same question. Find at least one hallucination. Document which tool was best and why.
Done when you have
- Tested 3 tools
- Found 1+ hallucination
- Documented comparison
02
Practitioner
Master
Prompting
Write prompts that get consistent, high-quality outputs every time
Core Topics
- Prompt engineering: roles, instructions, examples, formatting, tone, chain-of-thought
- Context engineering: giving AI the right background for relevant answers
- System prompts: setting persistent instructions and persona
- Daily AI practice: writing, summarizing, analyzing, brainstorming
Graduation Project
Your Custom AI Assistant
Write a system prompt from scratch. Give it a name and persona. Use it for a real task 5+ times. Refine based on outputs.
Done when you have
- Custom system prompt
- Used it 5+ times
- Refined it
- Documented final version
03
Builder
Ship Your
First App
Build and deploy working tools, no previous coding needed
Core Topics
- CLI basics: navigation, 10 commands you need
- IDE environments: VS Code, Cursor, Antigravity
- Vibe coding: using AI to write code you don't fully understand yet
- Claude Code basics: installing, running, structuring projects
- Full-stack concepts: Frontend · Backend · APIs · Databases
Graduation Project
Your First Personal App
Build a simple web app or automation. Solve a real, daily problem. Deploy it live. Use it yourself for 7 days.
Done when you have
- Working app/automation
- Deployed at live URL
- Solves real problem
- Used for 7 days
04
Architect
Build Systems,
Not Just Tools
Agents, automation, and systems that run themselves
Core Topics
- What is an agent? How agents differ from chatbots
- Skills → Sub-agents → Agent Teams progression
- Scheduled agents: automation that runs without manual triggering
- MCPs (Model Context Protocol): connecting AI to your data, tools, calendars
- RAG — Retrieval Augmented Generation: give AI access to your knowledge base
- HITL (Human-in-the-Loop): add review checkpoints
Graduation Project
Your Personal AI Agent
Build an agent that runs on a schedule and does something genuinely useful for you. Must include real schedule, 1+ HITL checkpoint, error handling. Run unattended for 2 weeks.
Done when you have
- Agent runs on schedule
- 1+ HITL checkpoint
- Error handling documented
- Runs unattended 2 weeks
05
Expert
Ship Real Things
for Real People
Full-stack AI products, launched and used
Core Topics
- Full-stack AI app development: frontend + backend + model + database
- Building for others: UX basics, onboarding, reliability
- API design basics: building your own endpoints
- What makes a good AI product: where AI adds genuine value
- Security: prompt injection, how to prevent hijacking
- Keeping current: how to stay ahead as the field moves
Graduation Project
Launch Something Real
Build and ship a complete project powered by AI. Live, accessible to anyone. Get at least 1 real user. When you have your first user — you're done. You're a builder now.
Done when you have
- Live product with real URL
- At least 1 real user
- Documented user feedback
- Portfolio-worthy