A 6-module, 34-lesson program personalised to your industry and graded by AI every step of the way. Built for the people most AI courses forget: non-developers who want to ship real work.
This entire program — platform, curriculum, grading engine, this page — was built by one person using Claude, scheduled tasks, and multi-agent workflows. The course teaches exactly what built it.
You either get generic "prompt engineering" content that teaches nothing specific, or developer-first deep dives that assume you already ship production software. Neither is written for the people who stand to benefit most.
"Here are 50 ChatGPT prompts" lists. Watchable. Forgettable. Nothing you learn survives contact with your actual job on Monday morning.
API keys, agents, function-calling, RAG. Incredible depth — wrong audience. If you don't already live in a terminal, you bounce in the first hour.
Real Claude work, tailored to your industry. Every deliverable is AI-graded against a 4-criteria rubric. Built to be finished, not just bought.
Most courses fail the same way: you finish week one feeling inspired, then drift. Every decision here is aimed at keeping you in — and making sure the work you do is the work you actually need at your desk.
Choose one of six industries at onboarding — agency, trades, health, finance, retail, or property. Every example, every prompt template, every deliverable is re-generated for your context. The same course, rebuilt for you.
Each deliverable is scored against four criteria — specificity, structure, constraint clarity, outcome focus — worth 25 points each. You pass at 70. You can resubmit. Nothing is "watched and ticked."
34 lessons. 20 hours. Progress visible end-to-end. A Core Capstone at the finish line. Every screen answers the same question: what's the next thing to do right now?
Not theory from 18 months ago. This course is rebuilt every release against the latest Claude capabilities — Artifacts, Projects, Computer Use. You learn what ships, not what used to.
Core covers everything a non-technical person needs to move from "curious about AI" to "working daily with Claude on real tasks." Each module ends with a graded deliverable you can use in your actual job.
What Claude actually is, how it differs from other assistants, and how to read its answers like a practitioner, not a tourist.
The four-part prompt framework. Role, task, constraints, format. Used correctly and repeatedly across every lesson from here on.
Email triage, meeting prep, writing support, content drafting, research. The actual tasks you'll use Claude for every week.
Stop starting from scratch every conversation. Build Projects with persistent instructions, attached knowledge, and reusable context.
Artifacts, Computer Use, file operations, image understanding. The new surface area of what Claude can actually do on your behalf.
A single, graded project that ties every lesson together. Ship a Claude workflow that solves a real problem in your actual job. Pass it to complete Core.
The grading engine reads your submission, runs it against a four-criteria rubric, and returns a score, a badge, and written feedback explaining exactly what would push you higher. Pass at 70. Resubmit anytime.
No subscriptions, no upsells, no "premium" tier hiding the good stuff. Core is the full program — every module, every lesson, every graded deliverable, forever.
Core opens 1 June 2026. Price locks at $297 for the launch cohort. Everything you need to go from curious to capable, graded every step of the way.
Claim founding seat — $297