AI · Marketplace
Case 01
Launched and scaled an AI-powered wedding marketplace
I used generative AI features to drive engagement and buzz, while measuring success by bookings, not attention.
30K+MAU
6 monthsto break-even
9+ years across Grab, Dell, Vault Platform, and a company I co‑founded. I make ideas tangible fast, judge them on evidence, and ship the ones that move the business.
I'm a Product Manager a GenAI Builder, a Co‑founder a Patent Author & someone who'd rather show you a prototype than a slide deck.
AI · Marketplace
Case 01
I used generative AI features to drive engagement and buzz, while measuring success by bookings, not attention.
30K+MAU
6 monthsto break-even
Consumer · Scale
Case 02
I owned the experience for the highest-value users (social sellers and corporate clients) across web and app.
5MMTU
+500Kmonthly orders
B2B · Enterprise
Case 03
I prioritized the enterprise roadmap, including an AI-powered hotline, and refined pricing and GTM to focus on enterprise sales.
100%revenue increase
Enterprise · Prototyping
Case 04
I led a 33-person team that built six proofs of concept in nine months, then decided which ones had earned a path to production.
$75Mrevenue impact
1patent granted
Medtech · Clinical
Case 05
I took a gamified VR training tool from concept to alpha in under six months, then validated it with surgeons and hospitals.
$1MSGD seed
$500KNUH project
Entrepreneurship · Hospitality · Operations
Case 06
I built and operate six vacation rentals, working closely with architects and construction teams to bring spaces to life, then rolling up my sleeves to keep the business running.
6properties
3 yearsSuperhost
Product in the age of AI
AI has made building cheap. What sets great PMs apart now is how fast they make ideas real, and how carefully they decide what's actually good. Here's what that looks like in my work.
From writing specs
I make ideas tangible before debating them. A working prototype answers questions in a day that a document takes weeks to settle.
Proof nPlasty's alpha went from zero to SGD $1M seed in under 6 months. At Dell, we built 6 proofs of concept in 9 months.
From acceptance criteria
For AI features, "done" means quality measured against real cases, including the edge cases where failure is costly. I define what good looks like with experts before we scale.
Proof At Vault, I planned how to adopt ML and manage its risks with the heads of Data Science and Security. At nPlasty, a $500K clinical study with NUH measured whether surgical performance actually improved.
From output
When almost anything can be built, choosing what's worth building is the job. I pick metrics that reflect real value and say no to work that only looks good.
Proof I made bookings, not followers, Weddi's north-star metric. At Grab, I focused on power-user segments, which led to profitability.
From features
AI products succeed when users trust them with real stakes. That means designing for privacy, human oversight, and clear failure modes from day one.
Proof I shipped a privacy feature in Dell Latitude with $75M revenue impact, and an AI-powered hotline in ethics and compliance.
My build loop
Weddi · Singapore
Vault Platform · UK (remote)
Grab · Singapore
Dell Technologies · Singapore
nPlasty · Singapore
SenseGraphics AB · Sweden
Quantic School of Business and Technology
National University of Singapore · Design-Centric Programme · NUS Overseas Colleges
English · French · Vietnamese