PRIVATE COHORT — CURRENTLY IN SESSION
The Upstream Shift
An eight-week rehearsal program that turns UI designers into product designers — on the work already on your plate.
8 weeks · Small cohort (3–5 designers) · Weekly 1-hour sessions · Homework is your actual job, done differently
The moment everything changed
You've probably had this moment already. A product manager shares a link in Slack. It's a working prototype — clickable, styled, not bad at all. They built it over lunch with an AI tool. No brief, no handoff, no Figma file. No you.
And somewhere underneath the polite "nice, looks great!" you typed back, a quieter thought: if they can do this part without me… what exactly is my part?
You're not imagining the shift. The professional whose job is receiving briefs and translating them into artifacts is being automated right now. Not in five years. This quarter.
But hold that fear next to this fact: at Gamma — one of the fastest-growing AI companies in the world — a quarter of the early team were product designers. Not cut. Over-hired. Because when AI commoditizes execution, someone still has to deeply understand the problem, the user, and the business — and that someone is worth more than ever.
Same industry. Same moment. Two completely different futures. The difference isn't talent. It's altitude — whether your value lives in the screens, or upstream of them, where the decisions get made and the tools can't reach.
You already know this stuff. That's not the problem.
You've read the articles. You know you should "ask why" and "understand the problem first" and "do research." You could give a decent talk about it.
And yet — when a request lands on Monday morning, you open Figma. When two PMs both say "urgent," you do both, in the order they arrived. When someone asks "what do you think we should do?", you answer with a question.
Research on product teams found that 98% of people asked to write opportunities write them as solutions. Staying in the problem space is the opposite of how our brains are wired. Nobody escapes that wiring by reading about it.
That's not a knowledge gap. It's a reps gap. Nobody becomes a confident public speaker by reading about speaking. You become one by speaking — in a room safe enough to be bad in, often enough that it stops being scary.
This program is that room. Almost no lectures. One concept a week, then you go apply it to the project already on your desk, then you come back and show what happened. Eight weeks, eight reps, on work that actually ships.
What actually happens to you in eight weeks
Week 1 — you stop taking briefs at face value. You learn the Ten-Minute Gate: four questions you answer before Figma opens. By Friday you'll have caught at least one request that was a solution wearing a problem's clothing.
Week 2 — you learn what your PM's week actually looks like. How PMs are measured, why briefs arrive solution-shaped, and the difference between pushback that earns respect and pushback that earns eye-rolls. You'll interview your own PM this week — and what you learn will surprise you.
Week 3 — you ship something ugly, on purpose. You learn to match the fidelity of your output to the certainty of your input — and you send a rough greyscale flow to a PM instead of a polished mockup. This is the scariest rep in the program. It's also the one where the ground shifts: nobody dies. The conversation gets better.
Week 4 — you replace opinion with evidence. Guided by a research practitioner, you run your first scrappy usability test: five people, twenty minutes, one task. Design debates run on two fuels — opinion or evidence. Opinion debates are won by seniority. Evidence debates are won by whoever watched a user last. This week, that's you.
Week 5 — you stop being a vending machine. You build a visible priority stack and run your week from it. When the second PM's "urgent" request lands, you have a sentence you've never used before: "Here's where it slots — and here's what moves if it jumps the queue."
Week 6 — you make AI your intern instead of your replacement. You take one step of your current project — research synthesis, divergent concepts, a testable prototype — and direct an AI tool through it. The designers who thrive aren't the ones defending the old process. They're the ones directing the new tools.
Week 7 — you learn to design the AI itself. Probabilistic products break the old UX rules, and most designers have never been taught the new ones. You'll map what data your AI feature actually has, and design the error states before the happy path. Roughly 85% of AI projects fail — almost never because the UI was ugly. They fail upstream. That's where you'll be standing.
Week 8 — you sell your work like a product person. You present one real project, end to end, on a narrative spine: problem → who and why → evidence → decision → ask. No screen tours. A story that ends with the room making a decision.
And here's the hidden curriculum: you present every single week. Five structured minutes, eight weeks running. By the capstone, presenting your thinking isn't a performance anymore. It's just how you talk.
The before and after
Opens Figma the day the brief lands → Runs four upstream questions first, every time.
Sees the PM as a source of requests → Knows how their PM is measured, and designs for that reality too.
Polishes everything, always → Ships rough work early and confidently.
Debates opinions → Brings evidence from real users.
First come, first served — for everyone → Runs a visible priority stack.
Asks "what do you think?" → Brings options and a recommendation.
Worries AI will replace them → Directs AI through the boring parts, and designs AI products others can't.
Presents screens → Sells decisions.
How it works
Eight weekly sessions, one hour each. Small cohort — three to five designers. Everyone presents every week.
Homework is your actual job, done differently. Zero synthetic exercises. Every assignment applies to a live project you're already responsible for. If you "don't have time for the homework," you've misunderstood it — the homework is your work.
One concept per week. A short pre-read, a 15-minute walkthrough, then reps. We never teach anything you won't use within seven days.
A shared board tracks your eight weeks. What you tried, what surprised you, what you'd do differently. By the end it's a visible record of your growth — the kind of evidence that promotion conversations are made of.
Who this is for
This is for you if: you came to product design through visual or graphic craft — and that craft is genuinely strong. You're respected, busy, and quietly aware that the job is changing under your feet. You're tired of being downstream of every decision that matters. And you're willing to be a beginner at something for eight weeks.
This is not for you if: you want a certificate more than a change in how you work. You can't commit one hour of session time plus applying the week's rep to your live work. Or you're looking for advanced Figma techniques — this program assumes your craft and builds around it.
The promise, plainly
Eight weeks from now, the people you work with will notice before you do. A PM will mention that your questions changed. Someone will ask you to weigh in earlier than they used to. You'll catch yourself saying "I recommend" without rehearsing it first.
You won't be a different person. You'll be the same designer — operating one altitude higher, where the tools can't reach and the decisions get made.
The screens were never the point. Come find out what is.
The first cohort is already running
The Upstream Shift is currently in a private pilot with a closed group of designers. I'm refining every week of the curriculum with real people on real projects before opening it up. If you want in when public enrollment opens, join the waitlist — waitlist members hear first and get priority on the limited seats.
Join the waitlist →
One email. Your name and why you're interested. That's it.