Products for governments, organisations and communities, including AI that has to earn people\'s trust. Each one opens as its own chapter: the problem, the people, the decisions and what changed.
No. 01 · Civic tech · AI · Media & journalism
People who witness floods, election irregularities or outbreaks often have no way to get what they saw to a newsroom, especially outside English and outside cities.
A WhatsApp reporting flow that works in any language, and a newsroom dashboard where AI confidence scores help journalists triage, while people still verify. Now live in a staggered launch.
No. 02 · GovTech · Public revenue
Cash-led, paper-based revenue collection made it hard for people to know what to pay, or to prove they had paid.
One portal to pay, verify and reconcile, with QR e-receipts for taxpayers and real-time dashboards and audit trails for agencies.
No. 03 · Enterprise · Information management
Offices ran on physical cabinets and shared drives. Files went missing, travelled over WhatsApp, or were lost to a single fire.
Live, with batch scanning and OCR to make migration realistic, and a Department → Cabinet → Folder structure people already understood. Stakeholders called it seamless. Next: AI-assisted classification built on that OCR foundation.
No. 04 · GovTech · Facility security
Facilities relied on paper visitor logs: illegible entries, no audit trail, and no way to know who was on the premises at any moment.
64 screens across 6 core flows, a full design system, and a check-in flow security officers can run without training.
No. 05 · Careers · Mobile app · Concept
Applying isn't always the hardest part. Keeping track of applications, documents, interviews and follow-ups across emails, notes and tabs can be.
One place to organise applications, compare opportunities, remember follow-ups and see progress, without making job tracking feel like another full-time job.
No two projects follow the same path. I move between research, design, building, and testing depending on what the problem needs. AI helps me work faster, but people, evidence, and judgment guide my decisions.
I start by questioning the problem, understanding who it affects, and identifying the constraints. I use research, conversations, data, and AI-assisted exploration to challenge assumptions before deciding what to build.
I turn insights into possible solutions, from rough sketches and user flows to interactive prototypes. I use Figma, AI, and code where useful to explore ideas quickly, without losing sight of usability and accessibility.
I validate whether the solution actually makes sense to people. Feedback, usability testing, accessibility reviews, and real-world constraints help me decide what to improve, remove, or rethink.
I work with engineers and stakeholders to turn designs into working products. From design systems and documentation to AI-assisted prototyping and implementation, I stay involved beyond handoff.
Going live doesn't mean the problem is solved. I look at adoption, feedback, usability, and outcomes to understand what's working. I also consider who might be excluded, what risks we've overlooked, and what needs to change.
AI supports my process, but it doesn't replace user insight, ethical judgment, or accountability.
next chapter: why any of this matters
