Yol Arkadaşım

About the app

Getting around the city — taxis and Uber especially — has stopped being affordable for most people when they travel alone. That is where Yol Arkadaşım came from: bringing together people heading the same way so they can share the same vehicle and split the fare fifty-fifty, or at whatever ratio the two sides agree on.

There were two ways to use it. Someone driving their own car could post a listing, find a travel companion and share the cost. Someone without a car could pick their destination, see the estimated fare for a taxi, Uber or TAG, then find people going along the same route and split the cost with them.

The most common question about the app was "does it work like Uber?" and the answer was no. The app did not carry passengers and did not sell transport services; it matched two people who were going to share a ride that was happening anyway. That distinction mattered both for how the product was positioned and for its legal side, so the interface spelled it out clearly.

The hard part technically was the matching. A listing carried more than just "from where to where": it also carried when (right now or a future date), which type of vehicle, and which people, matching which criteria, were allowed to see it. So the query had to work in both directions: the searcher's filters and the listing owner's visibility criteria had to hold at the same time. Add location proximity and a time window on top of that, and matching became far more than a simple "same destination" query.

On the pricing side the app was not a taxi meter; it calculated an estimated amount from distance and duration, and the listing owner stated how much of that amount they could cover as a lower–upper range. That way both sides knew who would pay what before getting into the vehicle — which is exactly where disputes tend to start in a shared ride.

Getting into the same car with a stranger was the real trust problem the product had to solve. That is why I put messaging and video calling before the ride; people could see who they were dealing with before heading to the meeting point. Letting the listing owner set gender and age range criteria came out of the same need. I still think this idea is a problem worth solving.

The app is not in the app stores right now — I put the project on hold for a while; I did not shut it down. I paused the release because I prioritised other products; the matching logic and the data model are still intact and I plan to keep developing it.