About — Chang Mou

Background

I'm based in San Francisco, hold a Master's in Interaction Design, and work on AI products. and hold a Master's in Interaction Design. I work on AI products, and specifically on the parts a model cannot take off your hands: deciding what to build, where a person stays in control, and what is actually good enough to ship.

I run my own research and build my own prototypes: in Figma when that is the right tool, in real code when a spec would lose too much, so the team gets a working build to extend, not a spec to translate.

I want to make people's lives better, and in this work that usually means taking something away rather than adding it. Those calls come before the data does, and taste is what you have instead. The model gives the hours back, and judgment is what they are for. Designers get more powerful in the AI era, not less.

Outside of work, I cook, walk the city on foot, read widely, vibe code some gimmicks, and spend time with friends. I listen to a lot of music, and more of my ideas come from it than from any designer. Nothing stays in its own field for long, and design is where it all connects.

How I use AI

AI's first pass drifts toward generic, so it never gets the last word. It runs the busywork; framing the problem, choosing the direction, and deciding what ships stay with me. The rule I work by is the one my products ship with: a human stays in the loop. AI accelerates everything and decides nothing. The judgment is the job.

One session, 29 July 2026 Claude Code, on a product's v2

Chang

just compare v2 and v1 in your opinion. if v1 is better why i do v2

Claude Code

Here's my actual opinion: v2 is the better product. Not the better-executed one today, v1 is more polished where it counts, but the better architecture. And architecture wins, because polish transfers and skeletons don't.

Show the rest of the sessionHide the rest of the session

Claude Code, continued

[three paragraphs of reasoning]

Notice what my strongest v1 arguments actually were: matrix width, decision gravity, re-entry after six months. Those are execution problems inside v2. They're not arguments for v1's architecture. The asymmetry is the answer: v2 has fixable weaknesses, v1 has a foundational one.

Chang

do not change the structure but check the latest design principles on agentic workflow design. what you suggest I can improve this project?

Claude Code

Research fleet is out, four parallel sweeps: lab guidance from NN/g, PAIR, HAX, Anthropic and OpenAI; teardowns of ChatGPT Agent, Devin, Manus, Cursor, Comet and Mariner; trust-and-oversight literature; consumer-delegation patterns like agentic checkout. When they report back I'll synthesize the findings against v2 and give you a prioritized list that leaves the three-pane structure alone.

Chang

do 2,1,5,7

Chang, forty minutes later

push to main as V2, do not touch v1

Shipped

The same morning, as homewise-v2.vercel.app. The opinion was the agent's; the structure, the shortlist and the ship call were mine.

Experience

MercorProduct Design Expert2026
PM AcceleratorProduct Designer2025 – 2026
UNICEFProduct Designer2025
LitespaceProduct Designer2023 – 2024
Asset LayerProduct Designer2021 – 2022
BC & Alberta Guide DogsProduct Designer2020

Education

MDes, Interaction DesignCalifornia College of the Arts · GPA 4.02025
Certificate, UX DesignDesignlab UX Academy2022
BSc, Interactive Arts & TechnologySimon Fraser University2021

Recognition

Red Dot Design AwardPollenNav, StoryBloom2025, 2026
iF Design AwardPollenNav2026
NY Product Design Awards, SilverPollenNav, StoryBloom2025, 2026
European Product Design AwardPollenNav2025
ULI Market Street Reimagined CompetitionExhibited at Ferry Building2025
Bernard Osher Foundation ScholarshipRecipient2024 – 2025
CCA IxD Departmental ScholarshipRecipient2024 – 2025
Wall of PortfoliosFeatured2025

People I've worked with

“Chang impressed our team at Litespace with his ability to conceptualize and implement complex design solutions like the Payments Dashboard and team collaboration tools. His dedication and user-centered approach made a significant impact on our product.”

Zeeshan Shahid

Co-founder, CTO, CPO at Litespace

“Chang’s strong work ethic and creativity stood out during his time at Asset Layer. He consistently adapted to team needs, delivered high-quality design work, and supported his colleagues with enthusiasm and collaboration.”

Dylan Straffon

Product Designer at Asset Layer

“Chang is not just a phenomenal designer; he is a true leader and invaluable strategic partner. He consistently delivered top-tier designs ahead of schedule while mentoring junior designers and elevating the people around him.”

Isabel Wei

PM at HouSmart (PM Accelerator)

I'm Chang Mou, a product designer working on what AI can't design yet.