Course · 63 lessons available
Build with AI, from scratch
The complete path from knowing nothing to building real products with AI — the same one I walked.
0/63 lessons
0 XP
Understand the terrain
FreeBefore you build, understand what you're playing with. No jargon.
What is AI? (seriously, no hype)
The one idea you need for everything else, explained like we're friends.
What an LLM Is and How It 'Thinks'
Tokens, context, and why it sometimes hallucinates. The bare minimum you need to know to use it 10x better.
The Model Landscape: Which to Choose for What
Claude, GPT, Gemini, open-source. When to use each one without getting overwhelmed.
The Builder's Mindset (You Don't Need to Be a Programmer)
The mindset shift that makes the difference between watching and building.
The tools
FreeChat, AI-powered IDEs and agents. Claude, Cursor, Codex, Antigravity — the model landscape and the art of the prompt.
Chat vs. AI-powered IDEs vs. Agents
The 3 ways to work with AI and when to use each one.
Claude, Cursor, Codex, Antigravity, and company
What each tool is, what it's good at, and when to choose it.
The 2026 Model Landscape: Which One to Build With
Anthropic, OpenAI, Google, xAI, and open-weight models. What each family does best and when to choose which.
Benchmarks: Which Ones Matter and Which Ones Are Marketing
How to read model metrics without getting hyped up. SWE-bench, contamination, and the only test that actually counts.
How to Structure Prompts That Actually Work
The core skill: going from mediocre answers to brilliant results.
Think through and scope your project with AI (before touching code)
The phase almost everyone skips — and the reason almost everyone fails. Research, scope, and turn your idea into a Master Prompt.
Set Up Your Workspace
How I get my environment ready to build fast.
Free vs. Paid Tools
How to choose: free tiers, when it's worth paying, and the traps of each option.
How to get unstuck: debugging with AI
The skill that separates those who move forward from those who give up: fixing errors.
How software works
FreeThe mental map: front, back, database, hosting — and how the pieces fit.
The Mental Map: Frontend, Backend, Database, Hosting
What each piece is and how they fit together, explained simply.
Yesterday's Stacks vs. Today's Stacks
Why building today is 10x faster than 5 years ago.
Programming Languages: When to Choose Which
JavaScript, Python, and company — no dogma, just clear judgment.
Choose your stack (the one I use and why)
Next.js + Firebase + Vertex + Stripe: my combo and why I use each piece.
Setup, terminal and commands
🔒 Full planCreate a project, organize it, and version it with Git and GitHub.
Creating a Project from Scratch
From an empty folder to a running project: Node, npm, and your first 'hello world'.
Organize Your Project (So You Don't Drown)
Where everything goes. A basic structure to keep you out of chaos.
The Terminal and Essential Commands
The commands I use every day (cd, ls, npm, git…) without fear.
Git: Painless Version Control
commit, branch, push. Why Git is a lifesaver (and how I use it).
GitHub and Working Safely
Repos, history, and never losing your work.
Backend and databases
🔒 Full planWhere your data lives: Firebase, Firestore, Storage and cloud functions.
What is a backend? (and when do you need one)
Servers, APIs, and data, simply explained.
Firebase in Depth (My Go-To Choice)
Auth, Firestore, Storage, and Functions: how I use them across all my products.
Supabase and the Alternatives
Supabase, Postgres, and other options — the big picture.
Which Database Should I Choose?
SQL vs. NoSQL, and how to choose for your project.
Cloud Functions: Serverless Logic
Run code on the server without managing servers. When and why to use them.
Containers and Docker (Without the Fear)
What a container is, why it matters, and when you actually need one.
From zero to your first product
🔒 Full planAuth, payments and professional design — from prototype to something people use.
From Idea to Prototype with AI
How I go from an idea to something that works in hours.
Users and Authentication
Login, sign-up, and accounts — what every product needs.
Getting paid: Payments with Stripe
From 'project' to 'business': subscriptions and payments.
Design That Looks Professional
How to make it look high-end without being a designer.
Create with AI: images, video, avatars and voice
🔒 Full planGenerate media with AI and plug it into your products — no designer or studio.
Generating Images with AI
From a prompt to a finished image: which model to use in 2026 and how to write great descriptions.
Generating Video with AI (Veo, Sora, Kling)
Generative video: what's possible today, which model to choose, costs, and real-world uses.
Integrating Generated Media into Your Product
How I connect image and video generation to real apps—securely and without going broke.
AI Avatars: Your Talking Clone (HeyGen)
Create videos with a talking avatar—even your own digital clone—for demos, content, and support.
AI Voice: Narration and Cloning (ElevenLabs)
TTS and voice cloning with ElevenLabs: give your content and product a voice, thoughtfully.
Services every product needs
🔒 Full planEmail, analytics, error tracking and integrations.
Transactional Emails with Resend
Get your app to send emails (welcome, password recovery, notifications) reliably and easily.
Analytics with PostHog
Understand what your users actually do, without losing your mind.
Error Monitoring with Sentry
Find out about bugs before your users start complaining.
How to Integrate Any Service
The pattern for connecting third-party APIs fearlessly (and securely).
AI agents and automation
🔒 Full planFrom asking AI for something to having AI run tasks on its own.
What is an AI agent?
The difference between asking AI to do something and having AI execute it on its own.
Build your first agent
From an idea to an agent that does tasks for you (like my Content Studio).
Composio: Connect Your Agents to Everything
Reddit, social media, Gmail, and 1,000+ tools without fighting with OAuth every time.
Security and best practices
🔒 Full planProtect your APIs, keys and users' data — and ship without fear.
How to Protect Your APIs and Keys
The #1 beginner mistake: exposing API keys. How to avoid it.
.env and Secrets: Where Everything Goes
Environment variables, what's public, and what should never be.
Security Rules (Firestore and Beyond)
Only the right user gets access to their stuff. Zero security holes.
The Most Common Security Mistakes
The checklist to sleep soundly before launching.
Ethics and Responsible Use of AI
Building with AI responsibly: transparency, bias, and not overpromising.
Basic Testing: Does It Really Work?
How to make sure something actually works before the world sees it.
Deploy: put your product on the internet
🔒 Full planVercel, domains, heavier backends and CI.
Deploying with Vercel
From your computer to the internet in minutes: the real-world workflow and environment variables in production.
Domains and DNS
Connect your own domain to your app and demystify how DNS works.
Cloud Run: For the Heavy Lifting
When Vercel isn't enough and you need a backend with real muscle (AI, PDFs, long-running processes).
Automation: Deploying Automatically
CI/CD without the jargon: every push publishes automatically, previews before launching, and scheduled tasks (crons).
Launch and get paid
🔒 Full planBuild in public, first users, marketing with AI and the economics of your product.
Build in public: building out in the open
How an open process becomes your best marketing strategy.
Your First Users and Feedback
How to get your first users and learn from them.
AI Marketing
How I automate my content and distribution with AI.
The economics of your product
What it really costs, how to set prices, and how to charge users without going broke.
The Legal Minimum (Without Being a Lawyer)
Privacy, terms, and just enough to operate with peace of mind — without formal legal advice.
My workflow — how I build
🔒 Full planThe real flow I used to build 13 products while operating by day.
My real workflow: a day of building
No theory: the exact cycle I repeat hundreds of times to go from an idea to working code.
How I Delegate to AI (and How I Verify Its Work)
The heart of the method: breaking down a problem for the AI, delegating it well, and never blindly trusting what it gives you back.
How I stay up to date without drowning
A new model or tool comes out every week. Here's how I filter the noise and learn only what my project needs, right when it needs it.
Your Next Steps (And Don't Go at It Alone)
The wrap-up: finish your project, make it your calling card, and join the community to keep growing.