Discovery Engine for Gaming Social Platform

A major challenge was managing millions of real-time updates from user profiles and game statistics. The platform needed to support instant search across high-volume data, without compromising performance.

To address performance and scalability, we implemented Apache Solr as the search engine and PostgreSQL as the main database. Solr’s indexing and query speed made it ideal for user discovery features, while PostgreSQL offered robust data integrity and relational capabilities.

The platform followed a microservices architecture:

• The User Profile Service managed gamer details like name, age, and game preferences.

• The Game Service handled player rankings, scores, and achievements.

To avoid complex data joins and improve search efficiency, we created a unified Solr collection that merged both services’ data into one searchable index.

Every time a user updated their profile or stats, the changes were stored in PostgreSQL and simultaneously indexed in Solr. This dual-write approach enabled powerful, fast user discovery across any combination of attributes.

Using Solr’s auto-commit feature and ZooKeeper, we ensured data consistency and high availability across the Solr cluster.

The result was a high-speed, scalable discovery engine capable of handling millions of profiles and updates. The system enabled real-time, multi-parameter searches with minimal latency earning high praise from the client and wide adoption by gamers around the world.

Our use of Solr, PostgreSQL, Flask, and ZooKeeper delivered a reliable, scalable discovery engine tailored for the gaming community. The single Solr collection approach minimized complexity and allowed instant retrieval of user and gameplay data creating meaningful connections between gamers based on their interests and performance.