We use it for messaging queue since we need a pipeline to process incoming product data. Besides, Spark has built-in analysis tool which is helpful.