Automation with EduCampus

EduCampus is a shared services company that provides business critical Information and Communications Technology (ICT) application services and Management Information Systems (MIS) to the Higher Education and Research sectors in Ireland.

EduCampus acts as a broker for clients by engaging with relevant service providers and suppliers in negotiating software licensing, application, hosting and support agreements in addition to providing implementation services and business operations support to the client base.

Introducing the Technological Universities Act

In March 2018, the Technological Universities Act was enacted to provide for the establishment of Technological Universities (TUs) in Ireland through the merger of two or more Institutes of Technology (IOTs). These mergers present several challenges for Institutes of Technology (IoTs) including the employee rehire process, where hundreds of employees need to be rehired on the TU’s HR and Payroll system (CoreHR). This is an extremely tedious, manual, and time-consuming task that must be completed by a strict deadline in parallel with other critical IoT merger and BAU activities. Each employee rehire takes four minutes to process manually, making this task a clear candidate for automation.

Version 1 has been developing Maximise: Automated Testing, a solution that automates quarterly regression testing on Oracle Fusion Cloud. Although a different service than CoreHR, Fusion Cloud is also a web-based service with similar use cases to the process of rehiring an employee.  It follows the same logic as a regression test that verifies that an employee can be hired.

Since the very outset of development, the Maximise: Automated Testing base framework has been designed to be transferable to other services as it was clear that there would be further use cases outside Fusion Cloud.

Solution

Defining the Rehire Process

Initially, an analysis of CoreHR was completed, to ensure there were no roadblocks to the automation such as third-party integrations that require the use of non-browser-based software. Next, the entire rehire process was verified to ensure that automation was feasible, and following this build the process was commenced, starting with the basic building blocks.

A modularised approach was taken to building the automation so that it will be possible in the future to easily swap steps in and out. This approach also speeds up debugging time as points of failure are easier to identify if and when they arise.

Since the information being used for this process is employee data, the development of the automation had to be done using only mock data and the process had to be developed in such a way as to allow EduCampus/IOTs/TUs to run the automation completely independently. The automation was driven by two data source files, one that would specify the URL and credentials to log in and the other to contain the rehire data itself.

Automation Process

As the automation process runs, it updates the data Excel with the result for each row. This is to ensure that if an error were to occur and the automation is halted, upon restart it will resume from the point of failure as previously successfully completed rows do not need to be rerun. There is also a backup log file that takes note of progress so that in the case of an unexpected problem e.g., network failure, there will be a mechanism to identify how many rehires were completed successfully. On failure, a screenshot is also taken to give a clear picture of what the screen looked like just as the error occurred.

Throughout development, Version 1 worked with EduCampus/IOTs/TUs as they provided feedback and refinements based on specific requirements. Some of these changes included:

• The automation will either stop on failure or can continue to process all rows in a loop.
• A test run was added to allow data to be checked, the test run does all steps but does not save the information. This will highlight any data discrepancies that may be present.
• The script can run on multiple environments with the only change being the login information

Outcomes

• The scripts were tested in a full UAT cycle, with the scripts it was possible to rehire all employees and run payrolls for all staff.
• Without the scripts, this would not have been possible to complete manually. In Live, the rehire scripts were run in test mode overnight with the results analysed in the morning, for the failures some required data changes to be made. The following night the scripts were run in save mode.
• For the initial project, there were two categories of staff on each site so overnight the scripts were run on multiple virtual servers and all were completed overnight. To rehire these staff manually in live would have required 14 days of effort.
• The scripts will be reused for future TUs when going through the designation/merger processes. Based on future TUs this process is estimated to save 120 days of manual effort.