This session is about achieving low latency decision making using Hazelcast and Iceberg. A new open source connector will be introduced. The current state of the integration will be reviewed and participants will learn how the combination of Iceberg and Hazelcast is relevant to low-latency ML Ops. The session will illustrate how to leverage the stream processing capabilities of Hazelcast to create reverse ETL pipelines that yield up-to-the-second features and how extremely high throughput, low latency scoring can be achieved by storing those features in Hazelcast distributed in-memory storage. If you need systems that react quickly to fast moving data this session will be of interest to you.