Data Integration: A Case Study
Companies integrated their data to gain business insights and eliminate risks. While individual databases can answer specific questions and show the state of a business unit, this is sometimes insufficient to make an enterprise-wide decision. The integration of all the enterprise data allows a complete depiction of the current state of the enterprise. Each case is unique, but the integration of data must provide cost-effective extraction and make transformation and flow of data between sources throughout the organization similar. When integrating data, you must pay careful attention to completeness, latency, accuracy, management, and ownership.
According to the case study, Teradata needs to prioritize its research and development commitments. In order to do this they must have all the information available and accurate. This requires integrating data from many sources both internal and external.
For this Discussion, consider and answer each of the following questions:
- Into which logical subject areas should the data be grouped? Consider the subject areas (customer, product, order) used for the Wheeler case described in your reading and determine a similar set of subject areas required for Teradata.
- Can your subject areas be considered independent components, thus allowing a single process for data integration to be defined? That is, if you define an ETL process for one subject area, can it be used for all?
- Is the data clean? At what point will you validate the quality of the data?
- How frequently must the data be loaded into the data warehouse? Is it the same for each component?