From 10+ Seconds to Under One: Solving Their Slowest Operation

A personal finance marketplace faced significant planning delays with its Apollo Gateway as complex queries took more than 10 seconds to process, leading to traffic congestion and an adverse impact on customer experience. To resolve this, the company transitioned to WunderGraph Cosmo, which resulted in a substantial reduction in planning latency by approximately 90–95%, making operations 12–18 times faster. This migration allowed the team to eliminate fragile prewarm scripts and stabilize performance, enabling safer upgrades and further growth. The phased rollout ensured customer traffic remained protected, while the implementation of Cosmo's Cache Warmer effectively preplanned slow queries, eliminating cold-start issues. As a result, the team achieved stable performance, routine supergraph updates, and predictable horizontal scaling, creating a more robust platform capable of handling increased demand without the need for constant patches or workarounds. This case highlights how addressing gateway inefficiencies can lead to significant operational improvements and supports sustainable growth for large-scale platforms.