BBGO Wins New Customers with a Cost-Effective Ride-Hailing Platform

Once we had a clear picture of the workflows, we explored the competitors like Uber, Uklon, and Bolt to identify must-have features and opportunities to improve the user experience.

The app for riders had to be intuitive and easy to navigate. So we prioritized core functions, like quick registration, instant ride selection, and flexible payment options. We also added extras like child seats and pet transportation for added convenience. This way, we didn’t confuse a customer with functions people rarely need.

For drivers, ease of use was also a priority. We added online registration to simplify onboarding, plus automatic order search and a built-in map showing accurate rider locations.

In just a month, we had key UI components ready – clear, intuitive, and mobile-first. From there, we transitioned into backend development.

Accurate location tracking is key for any ride-hailing app. We integrated Google Maps to pinpoint passenger locations and track drivers in real time. To keep it up to date, we used an XML live map that feeds real-time data from the backend, like driver positions and vehicle availability, directly into the map interface. This way, both passengers and drivers always see the most current info on their screens.

To match riders with drivers, we developed an algorithm that starts the moment a pickup is confirmed. When a rider sets the pickup point, the system searches for available drivers within 500 meters. The driver who has been idle the longest gets the request first. If no match is found, the search radius expands step-by-step to 1 km, 1.5 km, and 2 km.

We also added an algorithm where, if a driver is finishing a trip near a new pickup point, they’re offered the next ride immediately. It keeps wait times short for passengers and helps drivers stay busy and earn more.

To give passengers more convenience, we integrated multiple payment methods, including bank cards, cash, bonuses, and the Way4Pay payment service. Way4Pay also supports Google Pay and Apple Pay, allowing for seamless mobile payments right from the app.

Using the GeoJSON format, we mapped the city and its neighborhoods into zones (hexagons). We developed algorithms that automatically adjust prices based on activity in specific zones. For example, if more than 10 ride requests are made within a five-minute window in a particular hexagon, the price coefficient might increase from 1.0 to X2.5.

In addition to these automatic rules, administrators can create and manage manual pricing rules through the admin panel. For instance, on Wednesday at 3:00 PM, when there's heavy rain, an admin can manually set a rule from 3:00 PM to 4:00 PM with a price coefficient of X5. Once the time window ends, fares return to the normal rate.

At first, we developed the infrastructure on Google Cloud Platform, using Kubernetes and Helm. We also implemented a complete CI/CD pipeline, which allowed us to automatically build, test, and deploy new code with minimal manual intervention. This setup helped us build a scalable solution that could easily handle peak demand periods, like New Year’s Eve.

When the pandemic hit, the client began raising concerns about the high costs of running on Google Cloud. So we decided to migrate to Hetzner Cloud, a more cost-effective alternative. As the app infrastructure was designed to be cloud-agnostic from the start, using Kubernetes and containerization, the transition between cloud providers was fast, seamless, and disruption-free.