The monitoring of loading processes in data warehousing often poses a particular challenge. Even if all processes seem to run through without mistakes, this does not mean that the data is available in the required quality. In particular, non-obvious errors necessitate regular quantitative and qualitative checks of the data, which are time-consuming and must be carried out manually.
Based on the history, a model is created from which the data warehouse system learns how change results can be considered normal in the data - due to daily loading processes. On this basis, corridors can then be determined regarding the number of changes, but also regarding content elements, such as outliers in key figures or amounts of certain aggregation levels. If these corridors are then exceeded or undercut, the system reports this and a targeted error analysis is possible.
By using the biX AI Tools, the solution approach can be fully integrated into your SAP system.
Monitoring expenses can be significantly reduced, as regular monitoring is carried out by the system itself. In addition, deviations in both quantitative and qualitative aspects can be detected and analysed earlier. In the long run, this leads to a higher overall quality of the system.