A Taiwan-based social app that rewards showing up together. Users scan each other's QR at meetups to mint Friend Coins, redeemable at partner cafés, gyms and event spaces.
Unit of value: real-world presence — not matches or messages. Different from Meetup (event-led) and dating apps (1:1 swipe). MVP: one QR meetup loop + merchant redemption.
Real-world presence — valued, rewarded, habitual. The foundation of social life, not a novelty.
Social and dating apps optimise for what they can measure — matches, messages, time-in-app. None of those guarantee what users actually want: a real meeting in real life.
If the product can't make the offline meetup easier and more rewarding than the chat, it has already lost.
Audience
Teens & young adults in Taiwan — endless swipes, dying group chats, plans that never happen.
Existing alternatives
Meetup feels too event-led. Dating apps are 1:1 and transactional. Neither rewards just showing up.
Open scope
No prescribed goals. Users pick any activity — friends, dates, families, communities all welcome.
The original vision was untestable with our team. We reframed the core insight and cut scope to a single QR meetup loop we could ship and learn from.
V0 — Vision
Slow AI Dating
One AI match at a time, GPS proximity, in-app events. 6+ months to first ship.
Reframe
Meetups, not Matches
The real problem isn't matching — it's plans that never happen. Reward the meetup itself.
MVP — Shipped
QR + Friend Coins
QR meetup confirmation → redeemable credits at partner merchants. Testable in weeks.
Constraints that shaped every decision
Engineering capacity
2 part-time engineers. Anything beyond a few sprints had to be deferred.
No real user data
Pre-launch. Decisions reasoned from first principles, not analytics.
Shifting scope
Founder vision moved between dating, events, and rewards.
Trust & privacy
Anything resembling surveillance would kill teen adoption.
Product Workflow
Before designing individual screens, I mapped WiiTH as one connected journey — from onboarding and meeting someone in real life to friendships, rewards and redemption. This exposed missing states early and gave every screen a clear purpose within the wider product.
Design System
As WiiTH grew across onboarding, Meet Up, friends, shops and rewards, I built a shared component system so new flows could reuse the same interaction patterns instead of being designed screen by screen. Components were defined around meaningful states, from input feedback to the full Meet Up journey, keeping the experience consistent as the product expanded.
One system, multiple journeys.
Prove a small tangible reward is enough to make young people meet up — and that local merchants will pay for the resulting foot traffic.
Everything else cut.
01
Two users scan each other → both receive Friend Coins. Mutual scan = mutual intent. Zero background tracking.
Trigger
Mutual QR scan
Reward
Friend Coins
Privacy
No background tracking
02
Coins redeem at partner cafés, gyms, events — anchoring digital action to physical value.
03
Observe behaviour first. Validate demand before adding enforcement.
Phase 2 — Frictionless Presence
Documented, not dropped
GPS Passive Detection
Opt-in proximity + co-location duration → auto Friend Coins.
Event Infrastructure
Group QR, multi-participant invites, auto reward distribution.
Subscription Rituals
Recurring meetups (weekly dinners, fitness) with streaks & multipliers.
Five points where ambition collided with reality. One was load-bearing — the rest, supporting trade-offs.
Problem
Auto-detecting meetups via background GPS would drain battery, hit OS privacy walls, fail indoors, and be easy to spoof.
Decision
Ship a mutual QR scan handshake. Defer passive detection to Phase 2.
Why
Mutual scan = mutual intent. Zero background tracking. Familiar pattern, zero learning curve.
Trade-off
Adds one deliberate action — but buys trust, reliability, and a testable reward loop.
Problem
Two users could fake-scan to farm Friend Coins.
Decision
Observe first. Defer enforcement.
Why
If it's worth hacking, it's worth building. Coins only redeem at merchants — even abuse drives foot traffic.
Trade-off
Accepting some early gaming in exchange for shipping speed and real signal.
Problem
A long sign-up form killed conversion.
Decision
One input per screen, auto-advance.
Why
Form felt like a conversation, not a wall. Reduced perceived effort, lowered drop-off.
Trade-off
More screens to build, but dramatically higher completion.
Problem
Two-sided cold start: rewards app is worthless if no merchants nearby.
Decision
Onboard cafés, gyms and venues before users.
Why
First scan unlocks something redeemable nearby — instant payoff, no empty-app problem.
Trade-off
Slower public launch in exchange for a working reward loop on day one.
Problem
Proximity-based social features can easily feel like tracking.
Decision
No background location. Presence is only used when the user actively enters the Meet Up experience.
Why
WiiTH should help people find each other intentionally, not monitor where they go.
Trade-off
Less passive discovery, but clearer consent and greater user control.
MVP launched in Taiwan with a small set of partner merchants. Without mature analytics, the honest measure of impact is what changed for the team — and what it unlocked next.
Scope clarity
Cut the buildable surface from a dating + AI + GPS platform to a single QR meetup loop.
Faster delivery
Onboarding and QR flow spec'd as small, independent pieces — engineers shipped in parallel.
Investor traction
A working MVP turned investor conversations from concept pitch into product walkthrough.
Merchant pipeline
Anchoring rewards in redemption gave founders a concrete pitch: foot traffic.
Trust by default
QR over silent GPS pre-empted the privacy objections that derail teen-facing apps.
Honest roadmap
GPS, events and rituals weren't dropped — documented as Phase 2 with constraints to solve first.
The most useful thing design did on WiiTH wasn't the screens — it was turning a moving vision into a buildable product the team could agree on.
Startup product design is mostly subtraction under ambiguity. Three takeaways shaped how I now approach early-stage work.
01
Cut to the smallest testable claim.
"Slow dating with AI matching and GPS" isn't a hypothesis — it's four. Reframing it to "will a small reward make people meet up?" made the project shippable.
02
Constraints are the brief.
Part-time engineering and zero analytics weren't problems to design around — they were the brief. The QR-first MVP only makes sense when those constraints are taken seriously.
03
Trustworthy beats clever.
QR over invisible GPS wasn't the cleverer answer — it was the trustworthy one. In a teen-facing product, trust compounds faster than novelty.
Tensions held
Growth
vs
Control
Incentives
vs
Authenticity
Friction
vs
Verification
Vision
vs
Buildability
AI ProductIndependently designed and built an AI-first job application workspace. Candidate Profile anchors every output. AI tailors from the candidate's own record — not generic content. Supervised browser automation keeps the candidate in control throughout.
Graphic & UI/UXDesigned UI/UX for WUZO, a fintech platform featuring peer-to-peer payments, bill splitting, and QR code payments. Developed site maps, flowcharts for transaction history and KYC process, and built scalable design systems for consistency across platforms.
Brand IdentityCreated the full brand identity for PHYASE //, a Hong Kong physio and sport performance brand — from concept naming, tagline development (EN & Traditional Chinese), to logo construction, photography direction, and brand guidelines delivery.
