I am using Spark streaming for processing multiple CSV files. File validations changes often due to downstream systems. For this we decided to explore Apache Drools. Before