System One Engineering is a hands-on course about fast, measurable decisions: which team should handle this ticket, is this email spam, should this command be allowed, which way should the paddle move. You will learn the idea behind System 1 and System 2 thinking, how a System One model like Jev answers a narrow question with a calibrated probability, and how to design the whole decision around it: the state it sees, the questions it answers, the thresholds that turn answers into actions, and the fallbacks for when things go wrong.
You do not need to be a programmer. You work the way vibe coders work: you describe what you want to Muse Code, a coding assistant, then read, run and check what it builds. Every lesson is short, shows the real result, and keeps the evidence: requests, responses, logs, costs and honest limits.
Along the way you build real projects: a ticket router, an inbox triage system for a thousand emails, a cascade of two models, a review queue with a dashboard, a guard for an AI agent’s shell commands, a verifier for quotes, a real-time Pong player, a driving simulator with a safety layer, and a browser pilot. In the last section you prepare your own System One project for release with a launch checklist and monitoring.
