The data lake is powered by Azure Data Lake. Which is Hadoop and some other technologies. Adobe's Catalog is proprietary as well as some of the organization like data sets, batches and other pieces.
Query service is an interface to make *SQL-like queries. So, it's structured like sql calls but we have built in functions and libraries to help accomplish what some of Adobe's solutions do such as sessionization etc. The queries will query the data lake as well as profile service.