Problem Statement

• Delivering robust software requires testing an application with realistic data. For a software to be future proof, it must be tested on a high number of representative data sets to avoid possible future breakdowns and cost-intensive code corrections. Furthermore, the test data must cover lots of different test cases to improve the robustness of the software.
• Finding and adapting production data consumes a big portion of the developers’ time and always bears the risk of accidently spoiling sensitive personal data.
• Programmers are often confronted with poorly documented production systems where the relations between the data are not clear. This makes it difficult to understand the data model of the production system and therefore complicates further software development.

bankmark’s Solution

The Parallel Data Generation Framework (PDGF) enables software developers to test their programs with any amount of realistic data. The easy configurability and high speed of PDGF enables programmers to quickly adapt the test data set to software changes or to new structural findings within the production data. During the whole process, it is impossible to reveal sensitive personal data. Although PDGF‘s generated data look like the real production data, they still are synthetic in nature and do not have anything in common with the production data sets.

bankmark’s tools around PDGF allow scanning the production data to instantly get an initial schema configuration file matching the production data schema. This data description can be extended afterwards to improve the test coverage and to model more complex relationships between the production data which are not extractable from the database but are present within the data. This requires an in-depth investigation of the production data structure, which in the end leads to a better understanding of the production data organisation and finally a more realistic software test result.