The overall flow of an Agile BI methodology.
The Kimball Matrix: the dimension columns and the datamart rows
The concept of a ' Dry Spike'. Why Inception iterations (Dry Spikes) are not 'potentially releasable' (because ETL and schema design is avoided).
The importance of metadata to insulate the content and the user experience (which should not change significantly after the inception phase) from the underlying structures (which will change throughout the elaboration and construction process).
The importance of iteration: iterations for proof-of-concept and requirements, iterations to build one datamart at a time, etc
Discuss different ways datamart projects can relate to each other: sequential, overlapping (sashimi) or parallel.
Diagram of the overall process
Discuss doing inception and elaboration with open source and then switching to proprietary products for construction.