Case Study: Adobe achieves scalable, low-latency edge profile caching with DataStax

Adobe - Customer Case Study

Adobe Systems’ AudienceManager team needed a highly scalable, low‑latency data platform to power its Profile Cache Server (PCS), the distributed edge cache for real‑time user profiles. With about 6.4 billion profiles in the full store and 700–900 million kept on SSDs at the edge, the team required a solution that could handle very high read/write volumes, meet strict SLAs (around 12 ms for 95%+ of requests), and support cross‑colo replication and operational simplicity.

The team chose Apache Cassandra for its near‑linear scaling, tunable eventual consistency, and automatic cross‑data‑center replication, with DataStax improvements enhancing stability and operability. Running a 16‑node SSD cluster split between Los Angeles and New York, Adobe achieved consistent low‑latency performance, simplified replication and management, and the ability to quickly scale the service to support growth.

Adobe

David Weinstein

Director of Software Development