Adobe + Aerospike: Real-time Streaming Segmentation
summary
Adobe RT CDP replaces a Cosmos-style direct lookup with a decomposed, incremental state pipeline backed by Aerospike storage, loading up to 20 data sets per profile and resolving 5–50 identities before evaluating segments. The real-time path handles spikes of 100,000 events and evaluates roughly 10,000 queries per sandbox, reporting ~100x higher throughput (≈1,500 → ≈150k rps) and latency improvement (P95 ≈300ms → P99 ≈20ms); multimode tenancy and namespace isolation cut cluster count 5–10x and stabilise IO and ops.
AI-written summary · about
chapters (8)
- 0:00Introduction to RT CDP
- 3:43Segmentation requirements
- 6:30Query processing examples
- 9:15Workload characteristics
- 11:55Real-time system architecture
- 14:19Aerospike-based solution
- 18:16Infrastructure optimization
- 20:20Results and future outlook
related talks
33mUsing Aerospike as a Real-time System-of-Record for Adobe Experience Cloud Services on Azure
39mEdge Compute Infrastructure for Experience Business
32mReal time customer experience across industries
40mMastering small chunks of data in high frequency ad-tech systems
32mDistributed Personalization in Edge Computing Environment
48mRebuilding a RT Datastore: the design, deployment & performance story
48mThe real-time database to build your AI future
50mAerospike Technical Overview