Hi,
As in many companies, the company I am working with is considering implementing a data warehouse from scratch. The visualization tool of choice is QlikSense, TalendD and a few mySQL enterprise DBs for data warehousing.
As we all know data warehousing is a complex subject. I have reviewed the documents attached to this post.
I am wondering in terms of best practices, should ETL process be implemented within Qlik using scripting and therefore having a bloated visualisation tool with complex data models and data manipulation.
Alternatively, the visualisation tool is kept as lean as possible and all the data is pre-processed in advance, keeping QlikSense scripting and data modeling at a minimum. One school of thought is that the engine is optimised to handle large amounts of records and so this later becomes really complex to maintain and bloated with business logic.
An feedback from experts in this field is much appreciated.
