Case Study: KIXEYE achieves order-of-magnitude faster, scalable cloud analytics with Snowflake

Building a Cloud Data Warehouse at KIXEYE

KIXEYE, a studio that builds high‑fidelity mobile and online games, relied on event-driven analytics to run experiments and monetize gameplay across titles like VEGA Conflict and Battle Pirates. Their Kafka → Hadoop → Hive pipeline required heavy ETL to reshape JSON event logs into relational tables, created operational complexity and expense, introduced contention between teams, and caused long latencies (often more than two hours) before data was queryable.

KIXEYE moved its JSON event flow to Amazon S3 and deployed Snowflake as the central data warehouse, doing transformation and processing inside Snowflake using its native semi‑structured support and independent virtual warehouses. The change eliminated much of the infrastructure overhead, enabled concurrent workloads without contention, made data available an order of magnitude faster (processing up to ~1 million batches every few minutes), reduced costs through elastic, pay‑for‑use compute, and simplified operations so analysts can get to insights faster.

KIXEYE

Josh McDonaid

Director Of Analytics Engineering