Is it possible to create Spark job in Data Lake instead of a U-SQL job ?
Here are the options for big data services in Azure: