One developer, one market, no ads
Building the website and Android app for Kalimati, the market that sets vegetable prices for half of Kathmandu, on my own.
· 2 min read
Kalimati Fruits and Vegetables Market supplies 60 to 70 percent of the Kathmandu valley's produce. Its morning price board decides what a kilo of tomatoes costs in a huge number of shops that day. Traders, journalists, shoppers and economists all want those numbers.
I built the new website and its Android app on my own: design, backend, admin, content, deployment.
What it does
- Daily prices. Each morning staff enter the minimum, maximum and average price for every commodity, and it's public within minutes.
- Arrivals. How much of each vegetable, fruit and fish came in that day.
- History. Price and arrival trends over time, and side-by-side comparisons between dates.
- Traders. Dues and notices, without a trip to the office.
- A notice centre. Tenders, public bills, right-to-information reports, and daily pesticide residue test results.
- Nepali and English throughout.
- A newsletter that sends only one thing: the day's prices.
No ads, on purpose
The Android app is minimal and has no ads. Price information from a public market is something consumers have a right to, and an ad-stuffed app gets in the way of that. The app opens to the day's prices and doesn't ask for anything.
The admin is the real product
The public site is what people see. The admin is what decides whether the numbers go up at 7 a.m. or at noon. It had to work for staff who are not technical and who have a queue of traders in front of them:
- Role-based permissions, so each person sees only the parts they handle
- Two-factor sign-in, because it's a government site
- User impersonation, so I could see exactly what a staff member saw when something went wrong
- Sortable, searchable tables over everything
What I'd do differently
When it launched I wrote publicly that it delivered little in terms of looks and interaction, and I meant it. A solo build on a deadline gets the data right first. Since then, design has become half of what I care about. If I built it today, the price board would be the homepage, readable on a cheap phone in direct sunlight, and nothing else would compete with it.
I also wanted Nepal Bhasa alongside Nepali and English, and that still matters to me. Language is part of access.
See it: kalimatimarket.gov.np · the case study