James Dixon coined the term data lake in 2010 with a simple pitch: store everything raw, decide the schema later, skip the ...