How to build an 'Uber app' for bus ordering using open-source in just 1 month

Indonesia is home to over 270 million people. Ride-hailing is the largest segment of the mobility industry in the country. As we now know, its revenue hit $7.9 billion in 2019 and at the time, it wasn’t yet taken over by the big companies. bussr was fully aware it wouldn’t stay dormant for long.

Beating Uber, Go-Jek and Grab

It was obvious that the giants would not miss an opportunity to capture the growing Indonesian market, but as it goes with giants, they’re not the fastest creatures. In the beginning of 2019, bussr set itself a mission to beat the largest players like Uber, Go-Jek and Grab and conquer the market before them. With limited resources and time, an Indonesian bus-hailing startup needed to develop, test and release their product in the shortest time possible.

To help them with their endeavour, they picked Software Brothers because of our JavaScript expertise and experience building complex, scalable as well as cross-platform solutions — in this particular case mobile applications.

Google Assets Tracking meets AdminBro

From the very beginning, it was clear to us that the selected technologies needed to withstand the load of a large number of concurrent users in order to offer a reliable user experience.

From a business perspective, it was also important to create a back office system for the bussr C-level execs and employees that would allow them to monitor and manage the entirety of their operations.

Considering the time restraints, we decided to use our open source auto-generated Node.js admin panel — AdminBro — as a foundation for the backend and customize it in accordance with specific business requirements.

The key elements of the system were a map-based tool for assets and ongoing rides tracking (we picked Google Asset Tracking and implemented it into our AdminBro panel) and an advanced route creation and management system, synced with mobile apps which were being developed simultaneously.

MVP in 1 month

As a result, we managed to develop a working MVP, ready for closed testing in a real environment in just one month! The MVP consisted of a back office system and 2 x React Native Android applications (the platform was deliberately chosen due to the Android’s dominance of the local market) as we had to develop separate apps for drivers and riders.

The real life usage allowed us to continue the development while receiving crucial feedback from the bussr users, who were using the app every day in order to provide detailed data on the app’s performance.

As we proceeded to gain traction and were expanding the team to work on experimental solutions, like implementing a sonic-based ticket verification system, we also managed to create 2 x iOS apps, based on the shared React Native code, in just about 2 weeks.

Another example of a quickly delivered MVP can be found in one of my latest articles about CarFellows and their story of selling over 600 brand-new cars during the pandemic. If you’d like to discuss your project, feel free to schedule a call or drop me a line at kris@softwarebrothers.co.