A guided learning path from absolute beginner to shipping production AI apps — with an optional track into FinTech & Quant engineering — the developer side of quant: build the backtesting, market-data & execution systems quants rely on (not PhD-level research). The Claude tutor explains, quizzes, and grills you on every concept — it just won't autocomplete your homework.
Built by someone who learned this stuff the hard way. Every shortcut, every dead end, every "wait, what does this even mean" moment — flattened into one curated path.
Claude walks you through every concept — powered by Sonnet 4.6 with extended thinking — asks pointed questions, and refuses to write your code. You learn by building, not by copy-pasting.
424 topics across 18 stages. Every one earned its place. No "watch this 17-hour course," no "here's 50 resources, good luck." Just the path.
You ship real things. The AI Teacher gives you build briefs, you make the calls, it reviews your decisions. The roadmap pushes you toward production AI apps from week one.
The new FinTech/Quant track extends the roadmap into the engineering behind money and markets — market-data pipelines, financial databases, honest backtesting, risk, and AI research tooling. You build toward the Quant Developer / Trading Systems Engineer role the same way you learned full-stack: real projects, in order, with an AI tutor that makes you defend every decision. It's optional and comes after the core path.
Python for financial data → SQL + financial databases → finance foundations → prediction & modeling → financial data engineering → quant developer (backtesting, risk, execution) → AI for financial analysis.
Full-Stack → FinTech Engineer → Financial Data Engineer → AI/FinTech Engineer → Quant Developer / Trading Systems Engineer. Each rung is a job, not just a topic.
Real costs, slippage, walk-forward, and leakage prevention — no fake edge. AI is used for research, screening, and risk review first, never naive auto-trading of real money.
The onboarding asks 3 quick questions and drops you in at the right stage. No "learn HTML basics" if you already shipped React. Skip what you know, dwell on what you don't.
Each stage has curated reading + a project. Check off topics as you go. The Notebook captures your wins and the gotchas you hit.
Stuck on a concept? The AI Teacher runs a 5-step lesson + quiz. Confident? Skip it. Either way, you keep moving toward shipping a real app.
9 core stages get you to a production AI app and a job; two electives and an optional 7-stage FinTech/Quant track take you further. Everything is free with an account — you pay only for the AI tutor.
Four mockups — illustrative, modeled on the live UI — to show the loop: pick a topic, run the lesson, take the quiz, save what you learned.
Last time we covered React Hooks. Today's lesson: Suspense & data fetching — the next required topic in your syllabus.
Alternates: error boundaries · useEffect (mastery 54%)
try {
const data = await fetch(url);
if (!data.ok) throw ...
}
const arr = []; guarantee about arr?
Mastery: 87% · 3 quizzes scored over the last 14 days
from-teacher lecture quiz
Create a free account and the whole roadmap — 424 topics, 150+ projects, community, glossary, and progress sync — is yours at no cost. Paid plans add the AI Teacher with a monthly AI budget billed at cost — we never mark up what the model charges. Save ~17% on any plan by paying annually.
No. That's the whole point. The tutor system prompt explicitly refuses to autocomplete homework — it walks you through the concept, asks you questions back, and pushes you to write the code yourself. You learn by struggling, not by pasting.
With a free account: the full roadmap content, every topic page, the glossary, 150+ build projects (easy → hard), progress sync, the Notebook, flashcards, and the skill graph. An account is required to access the material; the only thing behind Pro is the AI Teacher itself — model-powered lessons and quizzes.
Your plan includes a monthly AI budget billed at the model's exact cost — no markup. Lite includes $4/mo, Pro $12/mo, Power $35/mo. Need more? Upgrade to the next plan for a larger monthly budget. Unused budget doesn't roll over — it refreshes monthly.
One model, chosen deliberately: Claude Sonnet 4.6 with extended thinking. There's no model switcher and no other providers — a single, strong, reasoning-capable model keeps the tutoring consistent and high-quality. Your credits are billed at that model's real cost, with no markup.
No. The platform proxies through a server-side key for every supported model — you never see it, never pay the model provider directly. We handle the billing.
Each stage is a coherent chunk of skills: stage 0 sets up your machine; the 9 core stages take you to a production AI app and a job. Two electives (Mobile, Data Science / ML) and an optional 7-stage FinTech/Quant track (stages 11–17) go further. You don't have to start at 0 — onboarding drops you where you'll learn the most — and you don't have to finish in order, but the stages compound.
Bootcamps are linear and expensive. YouTube is unstructured and full of dead ends. This is the middle path: curated like a course, paced like self-study, with an actual tutor that holds you accountable.
Yes — one click in the Customer Portal (Stripe-managed). Your account drops to free, the AI Teacher locks, but you keep all your content, notes, and progress.
Free with an account. Pay only when you want the AI tutor.