Experts don't adopt AI because it's accurate. They adopt it when they can verify it, change it, and put their name on the result.

Experts don’t adopt AI because it’s accurate. They adopt it when they can verify it, change it, and put their name on the result.

I design that layer.

Deployed at Gleneagles Hospital Hong Kong (part of IHH Healthcare, the world's second-largest healthcare group) and at N2 Clinic Hotel Chinzanso Tokyo.

Deployed at Gleneagles Hospital Hong Kong (part of IHH Healthcare, the world’s second-largest healthcare group) and at N2 Clinic Hotel Chinzanso Tokyo.

Selected work

The same problem shows up wherever someone has to stand behind an AI's output — a doctor, an analyst, a user deciding whether to trust an answer.

The same problem shows up wherever someone has to stand behind an AI’s output — a doctor, an analyst, a user deciding whether to trust an answer.

I solved it where it's hardest. Medicine is the extreme case: the stakes are highest, the regulation is heaviest, and the experts are the least willing to defer.

I solved it where it’s hardest. Medicine is the extreme case: the stakes are highest, the regulation is heaviest, and the experts are the least willing to defer.

Longevity.Omics

2025.3— 2025.7

One architecture, two extremes: a 500-bed teaching hospital and a boutique clinic, in two countries under two regulatory systems.

Runs without on-site support.

Genomic analysis a physician actually trusts

Genomic analysis takes hours to interpret. Physicians get 15 minutes per consultation. The platform closes that gap — with every recommendation traceable to the study behind it, and every plan editable before it reaches the patient.

Primary Designer

0→1 Product Design

HONG KONG & TOKYO

Longevidence

2025.12 — current

Two hours to build both prototypes. One week to test 50+ physicians. 85%+ chose one architecture.

Three days later, a $200B incumbent shipped that exact architecture. We'd already avoided competing with them head-on.

An AI product physicians helped define, then chose to trust

Engineering and the product docs were building toward two different categories, and nobody had noticed. I let physicians settle it by using both. Their preference wasn’t about features — they wanted to validate an AI’s synthesis, not drive the research themselves. That answer set the product form, and the trust architecture followed: every recommendation traceable from composite score to scoring method to the primary study behind it.

PRODUCT STRATEGY

PROTOTYPE-DRIVEN DECISION

FROM DESIGNER TO FUNDRAISING STRATEGIST

About

Hi, I'm Shuixin Wang, a product designer who takes AI products from undefined to deployed.

I don't wait to be handed a problem. At QuantumLife I was hired to design interfaces; within the first week I'd flagged that the team was building two different products without knowing it. Settling that question — and the fundraising narrative that followed from it — became part of the job, because I treat whatever company I'm in as if it were mine.

I build with AI, not just for it. Outside client work I use Claude Code to ship working software — research tools I use daily — without an engineering team. When a project drops me into a field I don't know, I work through the literature with AI overnight and come back with better questions for the experts.

Before this: MFA in Design & Technology at Parsons, BFA at RISD. Bilingual English and Chinese, based in New York.

Hi, I’m Shuixin Wang, a product designer who takes AI products from undefined to deployed.

I don’t wait to be handed a problem. At QuantumLife I was hired to design interfaces; within the first week I’d flagged that the team was building two different products without knowing it. Settling that question — and the fundraising narrative that followed from it — became part of the job, because I treat whatever company I’m in as if it were mine.

I build with AI, not just for it. Outside client work I use Claude Code to ship working software — research tools I use daily — without an engineering team. When a project drops me into a field I don’t know, I work through the literature with AI overnight and come back with better questions for the experts.

Before this: MFA in Design & Technology at Parsons, BFA at RISD. Bilingual English and Chinese, based in New York.