Most large organizations have multiple departments which have a key need for trend analysis across different dimensions and periods for their business process. The traditional technique for achieving this is to invest in a Data
Ware House (DWH), which is built periodically, generally on a month-on-month basis. It takes from hours to a couple of days completing the cycle and then reports are generated which show an outdated snapshot of the business!
Also, with Data Ware House comes the need for enormous infrastructure and high maintenance.
However, the question is - do we really need to analyze the behaviors across all attributes or would the trends across certain focused dimensions suffice?
As demonstarted in the diagram below, data is closely factored in a way that reduces data components and transforms for further analysis. Identifying strongly interacting attributes in constructing models thereby
increases the accuracy of predictions and dramatically decreases computation time. This also helps in reducing data calculations from 37 points to just 7 points.
QuickStart is our new avatar of Data Ware House - we focus specifically on key needs for analysis. Here are some benefits:
Lower cost of ownership
Lower cost of maintenance
Lower computation time
Can be computed frequently (like at EOD), since only focused attributes are being built