The intended use is to be able to use things like Hadoop or other tabular text processing systems to glean information from enormous amounts of data, then once valuable insights are found, use the Data Lake source to process it into a form suitable for a data mart, or preferably, a data warehouse.