Case Study: MixRank achieves faster performance and major cost savings with Citus Data

MixRank cuts hardware budget 50x with Citus Data

MixRank, a customer discovery platform, faced the challenge of scaling their PostgreSQL database to handle tens of terabytes of time-series data for near real-time analytics without the prohibitive cost of an in-memory database solution. They needed to distribute data across a cluster, which was not possible with their vanilla PostgreSQL setup, creating a looming scalability concern.

By implementing the Citus database, an open source extension to PostgreSQL, MixRank distributed their data across a 20-node cluster with SSDs. This solution from Citus Data delivered phenomenal I/O parallelism and throughput, allowing MixRank to process 3 petabytes of data monthly. The results included customers seeing significantly fresher data, a competitive edge, and a hardware budget estimated to be 50-100 times smaller than an in-memory alternative, dramatically increasing performance while cutting costs.

MixRank