Supply Chain Management

Customer Relations Management

Business Intelligence and Scorecards

Mobility Solutions

The first thing to do when dealing with a project to implement a business intelligence solution is to bring business and transactional information from the original systems (ERP, CRM, HR legacy systems, spreadsheets, other diverse file formats, unstructured information, etc.) to the destination system (data warehouse).

The mentioned process of information transfer always consists of 3 essential and inevitable stages:

Extraction - It comes to defining how information is collected from source systems . The extraction can be done in many ways, such as insisting on one's own DB source, executing data selection routines or generating data files to be imported later in the DW system. The data extraction processes can get a handle on a very large volumes of information (for example, when trying to transfer historical data for the last 5 years of activity), so it is usual to program extraction processes in time slots with low or no activity in the source systems.

Transformation - The objective pursued by the task of transforming the source information is to consistenly consolidate different sources. The transformation is due usually to a business rules established for this purpose. In the process of transformation it is possible to influence both the values of future analysis dimensions (characteristics) and the ratios (numerical values). In the last case is not unusual to find mathematical computing operations, which can be simple or reach an important level of complexity.

Load - The ETL process ends with the load of information already processed according to the relevant business rules in the destination datawarehouse system. Inherently to the load process, data storing is produced data storage system in the BBDD DW system, which will have been previously designed to optimally accommodate the transferred information (multidimensional structure cubes, star structure, etc.)..

An ETL process can also be used as systems integration tool , because it consolidates information from different heterogeneous origin sources, in a single destination application. To make this integration optimal, the transformation of information must carried out uniquely and pursuing the standardization and homogenization of the data. Therefore, all ETL process carries implicit optimization tasks of data, deduplication from them and elimination of inconsistencies, obsolescences and unnecessary redundancy.

If you want more information about how in Libertis Solutions we can provide guidance on how best to undertake an ETL project within your company, do not hesitate to contact us (info@libertis-solutions.com). We will be glad to assist you and provide accurate information so you can make the best decision for your business.

   
   
  More information in the area of downloads >>  
Legal Info