Operational BI tools

A beautiful information environment in itself is of little use if you do not have the practical options to access the data. There will therefore be a need for a series of working instruments or tools that take on this task and guide their users in the information gathering process.

Fixed reports

In an operational environment, applications are mainly used to program so-called fixed reports . A commonly used application is Crystal Reports . The only interaction that is required from the business user is to start the report, which may or may not involve specifying parameters. The development is done by technically qualified people who know the database and, thanks to various forms of tuning, can guarantee the necessary performance and maintain the proper functioning of the database. The business user obviously loses flexibility, but within an operational environment it remains important to adhere to the principle that data input takes precedence over output. Its continuity must never be compromised. And the only way to continue to guarantee this is to keep control over the extractions within the technical environment. Some reports can be called from a screen application, as is possible with Oracle Reports , for example, while others are stand-alone applications. The basic principle remains that the underlying database is OLTP-based (see article Operational Querying). Today, the Business Intelligence market is dominated by analytical reporting tools such as Business Objects or Cognos . Their operation is based on the use of a so-called symantic layer or transparent layer that is woven over the database structure. This offers business users the opportunity to generate SQL statements themselves, without any programming knowledge. Although the principle itself is certainly encouraging, some caution is required within an operational environment. The layer – although secured – creates a direct path to the core of the database and, in the absence of good technical guidance, can have disastrous consequences. It is therefore not recommended to use such analysis tools in combination with OLTP databases. The control over performance and extraction within an operational system should always be in the hands of the programmers, system administrators and the DBA, all persons who have direct access to the database and therefore do not need a symantic layer to compile a query. In certain circumstances it may seem useful to provide the programmer with such a tool to simplify the generation of SQL statements and thus simplify programming, but then the technical logic is transferred to the layer and the need arises to use it every time. to continue development. Outside the operational environment, analysis tools are indispensable within the information environment.

Operational desktop tools

The principle of local data (see article Operational Querying) is a fact that cannot be ignored, whereby certain forms of information are not located centrally, but with the business users themselves. Due to its decentralized nature, its existence is often unknown to the employees of the IT department. It is therefore impossible for them to provide any form of permanent reporting. The business user will therefore need a so-called operational desktop tool with which he or she can achieve the necessary results himself. The most classic and still very popular desktop tool is Microsoft Excel . The principle of the spreadsheet is ideal for reporting because it also uses the table principle and can be regarded as an alternative database, as it were. Because the latest versions of such software also provide various options for interlinking spreadsheets and connecting to just about any type of database, they may at some point even be considered part of the enterprise’s network and database structure. The use of Excel has mainly retained its popularity due to the simplicity with which an acceptable result can be obtained. For example, generating graphs is as fast as with any specialized analysis tool. Normal use also does not require specialized knowledge, as Excel is part of the standard office environment. That is why the most complex analysis tools provide an option to export the extracted data to Excel. Some tools, such as Business Objects, even go so far as to make it possible to refresh the data directly from Excel, so that the tool itself can ultimately only be regarded as an engine that retrieves data from the database in a transparent manner. Excel also offers the great advantage that the data can be overwritten. And this is where the greatest risk lies. One may wonder how consistent the data presented is at that time. The sheer speculation that there is a risk of fraud or error can be a reason to centralize local data and run reporting via standard methods, using the full functionality of OLTP tools.