Case Study II - Payroll Development with Sensitive Data
Our client, an employer with 45,000 employees came to PSE Data Security with a specific data privacy requirement: the development phase of a PeopleSoft Global Payroll implementation using "real" data except for the data elements classified as sensitive. The sensitive data elements included salary plans, compensation values, dates of birth, pension plans, etc. All other data was to remain "real".
Our client had been searching, unsuccessfully, for a solution via major consultancies, software and hardware vendors with the following requirements:
- budget of maximum Eur 250K
- all online and batch processes / functions must run to success
- payroll results must be valid
- end to end scrambling process must complete in < 8 hours
- no additional hardware budget
- 9 week critical timeframe
Description of Solution
PSEscramble was the ideal solution as it anonymises selected data elements while guaranteeing continued and correct functionality of the application. PSEscramble allows processes such as Global Payroll calculations to run to success and for the results to be verifiable - a feature missing in other solutions.
A data scrambling model was produced enabling one button future scrambling runs (occasional maintencance of the scrambling model is necessary where migrations effect the objects in the model).
One of the major issues with an implmentation of this size is perfromance. PSE resources tuned the data scrambling process to the level determined by the client.
The performance of the payroll processes with scrambled data was the same as that with real data i.e. no performance impact once the data is scrambled.
The implementation of PSE scramble was well under budget and within the timeframe.