Message queues are the backbone of modern streaming systems, but when used as long-term data platforms, they introduce duplication, rising costs, and complex architectures. As real-time analytics demands increase, treating streams as transient messages is no longer sufficient.

This talk introduces Apache Fluss – a streaming storage system designed to make data queryable immediately upon arrival. By combining schema-first tables, native record mutation, columnar storage, and unified data tiering, Fluss enables fresh queries, simpler architectures, and reduced operational costs.

Through architectural comparisons and real-world use cases, the session shows how Fluss complements message queues and helps teams build faster, more sustainable real-time data platforms.