Jayasena explained: "The business problem basically was how we harmonise and standardise and expose data from markets and systems across disparate divisions. What we wanted was to get a cross-business, cross-regional view of data.
"We had to move from traditional ETL [Extract, Transform and Load] type modelling and relational data modelling to MarkLogic's enterprise scale NoSQL capabilities. For example, we don't know what the data is going to be used for, so we looked at a more flexible, schema on read approach."
Sign up for Computerworld eNewsletters.