Why I'm Building Tapak

For most of the last decade, I’ve been an advocate for sustainable urban mobility across Southeast Asia — championing car-lite cities, advising city governments, and helping bring shared-mobility services to Cyberjaya, Penang, Shah Alam, Bangkok, Chiang Mai and Singapore. My work has been about shaping how cities think about movement, and making the case for better streets to the people who plan them.

That work taught me something that has stuck: the hardest part of building a low-carbon city isn’t ambition — it’s the first read. Before anyone commissions a study, files a plan, or sits down with a planner, they need a rough, honest answer to a simple question: given where this site is, what does low-carbon even look like here?

Today that first read takes weeks. It shouldn’t.

What Tapak is

Tapak (Malay for site, or footprint) is a tool I’m building to give a preliminary low-carbon reading of a Malaysian site from something as simple as an uploaded map screenshot. You show it where you’re looking, confirm the locality, and it returns a short, LCCF-aligned read — the kind of directional answer that helps a developer, planner or community group decide whether an idea is worth taking further.

Two principles are non-negotiable, and they’re the reason I’m building it carefully rather than quickly:

• Preliminary, never official. Tapak is not an MGTC LCCF assessment, a certification, or a Diamond rating, and it never pretends to be. Every output is indicative and directional — a starting point for a conversation with a qualified planner, not a substitute for one.
• Reference-only, and it cites its sources. Tapak does not scrape or redistribute government planning data. It points to the gazetted plan, names it, and links you to PLANMalaysia’s I-Plan so you can verify everything against the official source yourself.

If that sounds like a lot of guardrails for something still this early, it is. Working with cities taught me that trust is the whole product. A tool that overstates what it knows is worse than no tool at all.

Why me, and why now

I’m not a planner and I’ve never claimed to be. What I am is someone who has spent years advocating for low-carbon cities — and who got tired of waiting for the tools to make the case, so I started building them myself.

Somewhere along the way, I taught myself to ship software. I now design and run production systems end-to-end — including an internal project-and-resource platform, the Core, that a real team uses every day, with an AI chat layer sitting on top of live operational data. Building that showed me exactly what today’s models can and can’t be trusted to do, and where a careful, well-bounded assistant genuinely earns its keep.

Tapak is where the two halves of my work finally meet: the advocate who has spent years making the case for low-carbon urban development in Malaysia, and the builder who knows how to put a vision model behind a hard verify-before-you-trust wall.

Where it’s at

Tapak is early. The landing page, the data model, the report template and the reliability tests exist; the core screenshot-to-report analysis is the piece I’m building now — in the open. I’ll write here as I go: what works, what I get wrong, and what I learn about building AI you can actually trust in a regulated, high-stakes domain.

If you work in planning, sustainability, or property in Malaysia and this resonates — whether you’d use it, want to pressure-test it, or think I’ve got something wrong — I’d genuinely like to hear from you.

Follow along as I build Tapak, or reach out at tapak@motiondigest.com.

— Johan Sammy