Decoupling Inference from State Updates in Low-Latency Feature Engines via Probabilistic Thinning (opens in new tab)

Streaming data systems increasingly underpin Machine Learning workflows that maintain large numbers of continuously updated aggregations. In production settings, each incoming event typically triggers read-modify-write operations to persistent storage, making high-frequency state updates a dominant source of latency, contention, and operational cost. In this work, we decouple inference from state persistence in streaming Machine Learning pipelin...