Quality Assurance Improvements

When gravity9 joined the project, much of the initial infrastructure and testing framework had already been established, including a BDD-based automation setup and manual regression testing processes. However, several limitations were slowing progress and reducing visibility across the team. Automation tests were stored in separate repositories per QA engineer, test execution was triggered manually, and some tests depended on others to run successfully. Regression testing only took place in the pre-production environment once per sprint, meaning bugs were often identified late in the development cycle. In addition, the team lacked clear processes for managing regression bugs and struggled with poorly defined user stories, including incomplete acceptance criteria and unclear design specifications.

To address these issues, the QA team introduced several process and technical improvements. These included creating a dedicated backlog mechanism for regression bugs to improve visibility, establishing clear “Definition of Ready” and “Definition of Done” criteria for user stories, and enabling automated nightly test runs in the test environment to detect issues earlier. The automation framework was also significantly optimized by consolidating repositories, removing test dependencies, improving session handling and wait logic, enhancing pipeline configuration, and strengthening security by encrypting credentials. As a result, the testing process became more efficient, reliable, and maintainable, contributing to higher product quality and faster development cycles. Over time, the impact was clear: the consultant QA engineer reported 15% more bugs, tested 40% more user stories, and created 440% more automation scripts, achieving over 90% automation test coverage.