Rockset

Rumble, an Israeli company, developed an application that converts user's steps into reward coins, which can be used to purchase unique products or services. The company initially used PostgreSQL to handle data comprising users’ step counts, with three different tables tracking daily, weekly, and monthly steps. However, as user growth started to increase, PostgreSQL performance began declining, especially during peak times. The application's responsiveness declined with around 20+ requests per second, as PostgreSQL was unable to maintain the latency required to serve the leaderboards and eventually ran out of CPU and memory. Rumble's users are goal-oriented and being able to instantaneously see their steps and purchase coupons encourages them to maintain their active lifestyles. Therefore, Rumble needed to deliver real-time, data-driven applications to meet those needs. They needed a database that could handle complex queries, scale easily as their number of users grew, handle high concurrency, maintain low-latency queries, and require low ops.