Implementing Data

I will be implementing a BI/Analytic for mid sized financial Institution about 20gb of data. main goal is to create Reports(bar-chart,pie-chart).I will break up the plan here.

*) Data Base will be 3 NF
*) creating a materialized view by making a profile inside a Database.
*) aggregating and indexing (joins) like creating a universe in BO.
*) creating Classes and table hierarchy for Multidimensional analysis inside the BI tool for different servers.


I welcome your suggestions

Buy one of Kimball's books.

I studied his books Data warehouse Toolkit, Data warehouse ETL Toolkit, Data warehouse Life cycle

Please guide me in this process

If the database you are using is the operational database, I would not implement materialized views. If you are creating a new database, why build it in 3NF if the intent is to use it for analytics?

Where in any of Kimball's books does he recommend building a normalized database as the basis for analytical reporting?

If it not with implement with Materialized view than ----> ?

Build a dimensional warehouse.