Status: 🟡 Draft — under review
Length: 8 weeks
Time needed: 5–8 hours a week
AI coding tools can be your greatest force multiplier — or the fastest path to an unmaintainable mess. The difference isn't the tool. It's engineering.
This course teaches you to use Claude Code the way a real engineer would: with discipline, with process, and with a clear understanding of what you're delegating and why. You'll learn to manage context windows, steer agents with persistent configuration, plan features that exceed a single session's capacity, build feedback loops that catch errors before they ship, and run autonomous workflows that produce trustworthy code while you're away from the keyboard.
The core argument is simple. There are two common errors developers make with AI coding tools: they delegate everything (and drown in spaghetti code they don't understand), or they delegate nothing (and burn out trying to hold it all in their heads). Neither is sustainable. What you need is the engineer's path — a confident, proactive middle ground built on timeless skills: communicating, anticipating, planning, decomposing, delegating, systematising, and parallelising. These aren't new ideas. They're engineering. AI just makes them non-negotiable.
By the end of the course, you'll have a reliable, repeatable workflow for building production-grade software with an AI coding agent — and a much deeper understanding of your own codebase in the process.
Prerequisites: Comfortable writing code in at least one language. Familiarity with Git and the command line. Experience working in an existing codebase (not just greenfield projects). No prior AI tool experience required.
What you'll need: A Claude Pro or Max subscription (Pro at $20/month works; Max 5x at $100/month gives longer autonomous sessions). A terminal and code editor you're comfortable with (VS Code, Cursor, or similar). Git installed and configured. A GitHub account for the later weeks.