Spec-Driven AI for UX Designers: from Vibe to AI Pilot
Alex M. Chong
UX Director, Practice Advisor, Coach
The UX career path from vibe prompter to piloting your own AI systems
Everyone is telling you to learn AI and figure it out, but nobody is leading you through it. Your company wants you to become AI native, but can't tell you what that means.
So we default to "prompting better", or "training" a chatbot on our design system, hoping it knows what we want. It comes back almost right, you fix the rest by hand, and none of it carries to the next session. Becoming AI native means thinking in systems: understanding how AI operates well enough to design how it works with you, what it surfaces and when.
Yes, in this course you'll learn how to build AI agents – but I'll be teaching so much more than that. You'll learn the mindset shifts that make you a systems thinker, ready for whatever AI becomes next.
Week 1 · Get Claude to hold your context and behave predictably, instead of drifting every session
Week 2 · Build an agent that refuses to leap to output, and walks a vague idea into a build plan
Weeks 3 & 4 · Make it run on your custom process rather than a generic one, and reusable for future projects
Your agent assists your thinking, never replaces your judgement. I'll take you from vibe prompting to building an AI aircraft you can fly.
What you’ll learn
Hands-on, week by week: from vibing with AI in chat to piloting a system you designed for yourself.
Build your agent in Claude, with a hands-on activity every lesson. Each week builds on the last week's system
Separate it into behaviour, what it knows about you, and the work, so you swap projects without rewriting who you are
Write gates that hold each stage until it's defined: the task named, the concept described, every interaction written down
Pilot it on a real project with gates that leave a file at every stage, then delete the chat and watch it hold
Diagnose why output drifts, and learn to anticipate how a session will behave
Spot the two dynamics that cap you: making wishing to a genie, and a Claude that only obeys
Keep what works, delete the chat, and start fresh without losing ground
Map your real process into named stages, each with what enters, what leaves, and the check that closes it
Instruct your agent until it builds its own calibration log, so it adjusts without you editing files
Write loading rules so each mode reads only what it needs, and stop it drifting as the system grows
Have the agent read its own logs back as a report, and see a working pattern you hadn't noticed
Package your system so it installs into the next project instead of starting over
Learn directly from Alex
Alex M. Chong
Director of UX Design, UX Practice Advisor & Career Coach
Who this course is for
Senior product and AI designers weighing a move toward leadership, who want to grow systems thinking and strategic influence.
Design managers and directors leading a team to AI native: turn the noise into a mapped workflow and a system that scales.
UX/UI designers and researchers who feel left behind, want to re-position their career and build AI agents to work the way they think.
What's included
Live sessions
Learn directly from Alex M. Chong in a real-time, interactive format.
Certificate of completion
Shareable on LinkedIn and for employer documentation.
Office hours for project feedback and critique
Dedicated office hours give you instructor and peer feedback on the project you carry through the course.
Community of peers
Stay accountable and share insights with like-minded professionals.
Alumni network
Stay connected to designers and researchers who all build and run their own AI systems.
Lifetime access
Go back to course content and recordings whenever you need to.
Discounts to future programs
Every future program you join after this one is discounted, for good.
Maven Guarantee
Your purchase is backed by the Maven Guarantee.
Free resource
Stop AI Vibing & Start Spec-Driven Designing with Claude
Break the habits that come from treating AI as a chatbot
Recognize the chatbot reflexes that keep you re-explaining yourself, and learn what to do instead.
Pilot AI instead of babysitting it
Specify the workflow and the parameters AI works inside, so you're directing the work rather than guessing if it lands.
Build AI its own workspace, not another chat window
Design an agent that carries your methodology — gating stages, producing documentation, driving the work forward.
Thrive as a designer while everyone becomes technicians
Turn systems thinking into your advantage: build AI systems you pilot, instead of chasing the next tool.
Frequently asked questions
Maven for Teams
Reimbursement
Get your company to pay
Everything L&D needs: email template, receipts, and certificate of completion.
Get reimbursedPrivate cohort
Run a cohort for your org
A dedicated cohort with a custom schedule and curriculum, tailored to your team.
Book a private cohort