"data Artisans Streaming Ledger [...] [extends] Apache Flink [...] to perform serializable transactions from multiple streams across shared tables and multiple rows
of each table." [1]<p>This sounds like a hybrid approach if you need streaming to handle the speed and volume of data, but still want to use ACID on a small subset of events.<p>LMAX [2] architecture may be another approach to consider for your application.<p>[1] <a href="https://data-artisans.com/wp-content/uploads/2018/08/2018-08-31-dA-Streaming-Ledger-whitepaper.pdf" rel="nofollow">https://data-artisans.com/wp-content/uploads/2018/08/2018-08...</a><p>[2] <a href="https://martinfowler.com/articles/lmax.html" rel="nofollow">https://martinfowler.com/articles/lmax.html</a>