I have spent most of my career in public service, working across strategy, operations, community engagement, organisational change and increasingly, technology.
This site is where I keep the things I’ve built, the work that shaped how I think, and what I’m exploring now.
Some projects begin with a problem I run into. Others begin with a question I want to test. They range from decision-support tools and searchable libraries to learning games and AI-enabled experiences. Taking an idea far enough to become usable forces me to think more clearly about the user, the trade-offs and what the technology can — and cannot — do.
Exploring how Primary 1 registration information can be organised around the decisions parents actually need to make.
A searchable collection of Singapore primary-school examination papers, built around discoverability and filtering.
Exploring where AI can genuinely help turn assessment information into more targeted practice, feedback and useful next steps.
Reimagining existing brand assets and training content as a more game-like learning journey instead of another page of material to click through.
A short learning experience designed to make vibe coding approachable to people who do not see themselves as coders.
An interactive pre-read that helps leaders grasp core Systems Leadership concepts before class, creating more time for application, discussion and judgement during the programme.
A Pair Chat design assistant built for a workshop I ran with public service leaders and managers, guiding them through the six questions behind a good custom chatbot — purpose, audience, experience, knowledge, guardrails and testing — before they write a single prompt.
Chatbot Architect only opens on Whole-of-Government (WOG) laptops, since it runs on Pair Chat. The guide above is open to everyone.
There was no single point where I decided to “move into technology”. It happened gradually. Each stage of my career exposed me to a different part of the system and pulled me closer to the next question.
My early roles in international operations, strategic planning, service excellence and organisation development gave me a lasting bias towards understanding the system before designing the intervention.
Working closely with constituency teams, community leaders and staff changed my perspective. A strategy can make sense at headquarters and still be difficult to apply in a particular operating context.
PRISM — a resident-data system rolled out across PA’s constituency network — reinforced that the challenge is not simply teaching people to use a system. It is helping the technology fit alongside new processes and ways of working.
A gamified digital learning application brought me through requirements, public-sector procurement, vendor discussions, development, beta testing, launch and post-launch adoption.
276 users took part in a three-month beta, generating more than 350 pieces of feedback.
More recently, I have used government digital platforms to automate work, built AI-enabled assistants for programmes, experimented with enterprise AI, and started building my own products through AI-assisted development.
One attendance and RSVP workflow reduced manual administrative work by 40%.
These are working principles rather than grand theories. They have simply proved useful across different roles and problems.
Technology matters when it changes the quality, speed or economics of solving something real.
Usage, support, behaviour, maintenance and improvement are part of the product.
I do not need to be the engineer in the room, but building makes me better at asking useful questions.
Shared platforms matter. So does giving teams safe ways to solve legitimate local needs.
A large part of my curiosity now sits around how AI, reusable platforms and new interaction patterns change what small teams can build — not only tools, but also better ways to learn, explore and make decisions.
I am interested in where agents genuinely remove friction, and where they simply add another layer of complexity.
The challenge is giving teams flexibility without creating a new generation of disconnected point solutions.
As AI becomes embedded in everyday work, the interesting design question is not only what can be automated, but what should not be.
I’m based in Singapore and have spent my career in public service since 2009. The work has moved through planning, service excellence, community-facing roles, capability development, change programmes and digital products.
“I tend to enjoy problems that sit between functions and do not have a neat owner.”
Different people can all be reasonable and still see the same problem differently. My usual instinct is to get close to the people doing the work, make the problem concrete, and find something we can test.
I’m still learning. A lot of what you see on this site exists because building something is often the fastest way for me to discover what I do not yet understand.
That is why ground engagement and direct user feedback matter to me.
The question is whether it makes the work better, simpler or more sustainable.
Pilots and prototypes are often more useful than pretending the answer is obvious at the start.
Adoption and support are part of the solution, not someone else’s problem.
I have kept this site intentionally selective. There is much more detail behind the work here — what worked, what did not, what I learnt and what I would do differently today.