Data Validation Procedures
BiG EVAL provides several out of the box procedures for data validation and plausibilization that are easy to use. If that's not enough, the scripting engine makes you capable of building your own algorithms.
Built in validation procedures
The built in validation and plausibilization procedures can be configured in the user interface and do not need any kind of programming skills.
Probes exactly equal
Compare an infinite amount of probes out of your systems.
E.g. Is the revenue of a specific product equal between the ERP, the CRM and the Data Warehouse? Even within an OLAP Cube.
Probes approximately equal
Perform plausibility checks on your data with a fuzzy approximation logic.
E.g. Does the monthly revenue of a product lie between +/- 2% of the previous months revenue?
Probes not equal
Does your systems have disparate KPI values?
Make a performance measurement for systems and database queries. Ensure that performance requirements get met every time. And let BiG EVAL send you a message when they appears to be a problem to respond proactively.
Monitor the schema of a database to be informed about schema changes. Do not miss any changes anymore. e.g. datatype changes, field-name changes, adding or removing any tables etc.
Checking data cell by cell by applying complex rules is the strength of BiG EVAL. Business Rules are based on an expression language and can be extended to complex scripts.
- Check data types of cell data.
- Check numeric or date ranges.
- Lookup information in foreign systems and check whether the cell data matches.
- Make complex calculations to validate or plausibilize cell data.
- Check data row by row.
- Run C# scripts as validation rules.
Outsourcing Validation Logic to the data base management system
When the preconfigured testmethods are not enough, you can build your own just by using the query language of your database systems (MS SQL, Oracle, Teradata etc.).
Scripting your own Validation Logic
And if that’s still not enough you can use C# as a scripting language. This opens much more possibilities than you can imagine.
- Execute a test multiple times. So you just have to build a test once and execute it with different parameters. e.g. you execute it for every sales-region and every product and so on.
- Take your test-parameters out of a database and apply them to testmethods.
- Prepare and cleanup your testing environment automatically.