About

I like building things that solve real problems.

I have always been most motivated by problems I care about and where I can see that the work has a real effect on people.

I started my career in policy and operations. During the pandemic, I worked on supply-chain resilience, where decisions often had to be made quickly with incomplete information. One thing I became interested in was whether we could use data and technology to make those decisions better. In some cases, we were able to reduce parts of disruption analysis from days to hours.

Since then, more and more of my work has involved building things from scratch. At Open Government Products, I worked on products like ReadLiao, which helped patients understand difficult letters and documents. I later helped set up the early learning products and programmes that became part of the Institute of Digital Government. Today, I build national anti-scam products that use AI and data to detect and disrupt scams across websites, mobile networks, social media and messaging platforms.

I have also learnt something about how I like to work. If I care about a problem, I tend to get quite deep into it. I am willing to read up, speak to people who know much more than I do, and get myself up to speed quickly if that is what it takes to make something useful.

The problems have been quite different. I have had to understand how people with very different levels of digital confidence learn, so we could build programmes for a workforce of more than 150,000 people. For Tamagotcha, I spoke to shelters and prospective pet owners to understand why people sometimes underestimate what owning a pet involves. For Hawkernomics, I spent time understanding the economics and everyday pressures behind running a hawker stall. In my current work, that means learning how scammers operate, how they persuade people, and how they use technology, so that we can find ways to make those operations harder.

Sometimes that means a product. Sometimes it means a policy, a programme or a different way of running something. I am not too precious about which.

The question I usually come back to is whether what we built actually made things better. Did it save someone time? Help them make a better decision? Make something safer? Help an organisation solve a problem it could not solve before?

That is also part of why I started DatastorySG. We have now worked with close to 900 students, helping them learn data and AI by building things themselves. As these tools become more powerful, I am increasingly interested not just in what I can build with them, but how more people can use them to turn their own ideas into something useful for others, while understanding how to use them responsibly.

Say hello

I’m always happy to meet people working on ambitious AI problems and technology with meaningful social impact. The easiest way to reach me is on LinkedIn. You can also explore what I build on GitHub or email me directly.