Challenge
The client faced significant capacity constraints, particularly in managing user login services for a platform serving tens of millions of monthly users, over 15 million daily active users, and millions of concurrent users during peak times—figures that double when users in China are included. To address these demands, the client sought to reduce operational overhead without adding personnel, aiming to enhance service efficiency and enable the team to manage new services without expanding. Additionally, they required a dedicated team to maintain related services, ensuring their optimal operation and upkeep.
Solution
We addressed these needs by standardizing service construction and operations, streamlining onboarding, and minimising learning curves for new team members. We updated services with modern libraries, migrating from proprietary frameworks to SpringBoot 3.2 and Java 21, reducing technical debt to enhance reliability and minimize maintenance. By replacing outdated solutions, we improved scalability and became the primary contact for resolving issues, proactively identifying and addressing common problems. Our improvements and thorough documentation ensured operational excellence, making integration and use easier for the client’s game teams. We standardized monitoring with the latest Datadog offerings, enhancing Dashboards and alerting, and introduced an automated deployment process via GitHub Actions.
Results
Building on our improvements, we focused on significant service enhancements, increasing the number of services under our management and bringing them to a production-ready state with minimal upkeep. Through in-depth domain analysis and strong client connections, we identified and implemented service improvements and infrastructure optimizations, resulting in substantial cost reductions. These included a ∼40% reduction in S3 storage costs by introducing a Redis caching layer and a ∼30% reduction in operational costs by downscaling over-provisioned services after extensive load testing. Our understanding of player traffic allowed us to renegotiate lower prices with the SMS/Phone ID API. Our holistic approach reduced monthly invoices by 40%, enhancing service reliability and operational efficiency.