Luiza Fontana

This project started as a proof of concept and ran over 1.5 million API tests in under a year. With only one other colleague working on it, we were responsible for requirements gathering, architectural design, development, implementation, and maintenance. The tool could be used during test analysis and regression testing and would save all the data for reporting on test execution, progress, and change over time in a MongoDB database. The initial clients of the tool were the Regulatory Testing Team, who used it to increase test coverage on tax tables and schedules and found over 38 defects that were successfully handled before going to production.

At a later stage, improvements were being made to run tests in parallel and remove any dependencies on the system under test, making it possible to use the same tool for different types of tests as well as many different teams, aiming at effectively unifying the test standards of the company.

In my first automation project, I helped leverage the underutilized potential of a tool already being used as a company standard. With some learning on my own time, I brought a use case to the project manager, who allowed me a couple of hours to complete a proof of concept.

When the automation was implemented, we could test a much larger combination of inputs, with the option to run a random few or all possible values of a given field. One notable example was ensuring that over 200 country codes were always working.

I built a GCP-based server and installed the test automation framework so that the tests run regularly on a schedule and report the results to a dashboard without human intervention. I used Python, bash scripting, and Crontab.