Alphaquant’s task was to audit existing Data Warehouses for Bertelsmann’s RTL TV Media Teleshopping. After analyzing the database, cubes and reports with it’s structures and processes we came up with many helpful recommendations regarding improvements for the data structure and performance. The return on invest for the audit compared to the outcome was huge.
Challenge
The reports generated out of a central data warehouse were running too slow. It took multidimensional cubes several cubes to open and reports too much time to finish. Some reports even hang up. As a consequence Alphaquant was tasked to audit the existing data warehouse and look for improvements. With our broad knowledge in database development, database adminstration and performance tuning at hand we are the right partner to audit any database or data warehouse system. The audit itself took only 2 weeks to finish.
Solution
Alphaquant met with the responsible team members and manager to define the scope of the audit of the existing data warehouse. After clarifying the scope in a kickoff workshop we examined the architecture on different levels. These levels were database design, ETL processes and multidimensional cube models on site. On each level we found several possible improvements. Some were quick wins and others required a lot of effort and general changes in the architecture. One main issue was that there were too many individual customers. On account of the fact that this granular level was not necessary we combined customers into buckets of certain demographics.
Result
At the end of the audit of the existing Data Warehouse we documented all proposed changes in a comprehensive word document for future reference. Lastly the customer’s internal development team finished implementing our recommendations the cube building time dropped by 50% and the reports sped up by 400%.
Technologies
Oracle, Cognos, DataStage, Transformer, Powerplay