Automating Regression Testing In a Domain Specific SaaS Setting

Abstract: This thesis aims to provide insights into how implementing an automated solution for regression testing in a domain-specific SaaS environment can address common regression testing challenges such as high cost or poor test result visualization. Additionally, it presents a practical demonstration of how to approach these challenges and the potential development of the solution. While previous studies on test automation often recommend adopting a third-party automation tool to tackle certain regression testing challenges, this thesis takes a different approach. We explore the feasibility of developing a test automation framework specifically tailored for a SaaS company operating in Sweden's insurance and pension industry. This industry is renowned for its complex infrastructure, and the company requires efficient testing to support frequent releases and accommodate its diverse range of customers. To identify domain-specific issues and answer the feasibility of automating regression testing in a domain-specific setting, this thesis begins with a literature review and interviews to identify generic and company-specific issues related to regression testing and test automation. After identifying these issues, the thesis proceeds through the development process of a test framework, test execution, and framework analysis. It also explores the growing interest in test automation and examines the ethical considerations associated with its implementation. Finally, the thesis provides a test automation framework development process starting with identifying the specific issues related to developing an automated solution for regression testing for a SaaS system, continuing with its development, and confirming the feasibility of creating such.