Behind the Build

GH Floods: How One Bad Commute Became a Crowdsourced Safety App

Stuck in traffic during heavy rain with no way to know which roads were flooded, we built GH Floods. It had 230 users in its first 24 hours.

July 15, 2026 4 min read

It started with a flooded road, an Uber, and no idea which route would actually get us to work. By the end of that day, the idea for GH Floods was already mapped out.

The commute that started it

It was a heavy rain morning in Accra. The kind where you know before you leave the house that something is going to be flooded — you just don't know what. The Uber driver was navigating blind, turning down streets only to find water halfway up the wheels, backing out, trying another route. We arrived late and spent the ride thinking about how there was no way to know which roads were passable.

People on Twitter/X were posting flood videos. WhatsApp groups were buzzing with warnings. But it was all scattered, unverified, and impossible to search by location. There was no map. No way to look up a specific route ahead of time.

The information existed — it was just stuck in individual phones instead of on a shared map where it could actually help people.

The concept: crowdsource the map

The idea was straightforward: build an app where drivers and commuters can flag flooded or blocked roads in real time, so the next person heading the same way can route around it. We wanted it genuinely frictionless — no account creation, no long forms. Tap the map where the flooding is, mark the severity, submit. Done in under ten seconds.

The harder problem was keeping the map accurate. Old flood reports that nobody has cleared are worse than no report at all — they erode trust in the app. So we built community confirm/clear voting (other users can mark a report as still active or cleared) and automatic expiry for reports that haven't been confirmed in a set window.

230
Users on Day One
0
Login required

230 users on day one

We launched during rainy season. Within 24 hours, 230 people had used it. That number says something important: the demand was already there. Nobody needed to be convinced flooding was a problem worth solving. The moment a tool existed for it, people used it.

The no-login approach turned out to be the right call — lowering friction meant people reported flooding in the moment, rather than deciding it was too much effort.

What we'd build next

A few things on the roadmap: push notifications for specific routes, integration with mapping apps, and a more sophisticated expiry algorithm that factors in weather data. The core product works today — GH Floods is live at ghfloods.augtechghsolutions.com. If you're in Accra during rainy season, keep it bookmarked.

Have a product idea?

We build software that solves real problems for real people. If you have a problem worth solving, let's talk.

Get in touch → ← Back to Blog