Alisa Hrustic

My name is Alisa Hrustic and I work as Test Manager for Alfa Laval. I hold a master’s degree from the Faculty of Electrical Engineering in Sarajevo.

I have 15 years of experience in the area of test automation combined with test management, with a main focus on technical improvements and building testing strategies to improve test team efficiency. In the last few years my focus has been shifted to exploring and learning about AI with special area of how AI is transforming software testing.

The main goal of this talk is to provide an engaging overview of how AI is transforming and enhancing software testing. In this case, AI leverages user experience data to generate new test cases and expand existing regression flows. As QA engineers, we know that even with well-designed and comprehensive tests, updates are not always frequent—due to factors such as time constraints or maintenance challenges. As a result, certain user flows, especially those without detailed requirements but commonly used in real scenarios, may remain uncovered. This gap often leads to user frustration, since it’s unexpected for real-world flows to fail when so many release tests have already been executed.

AI offers a way to address this challenge. The concept is to build a self-learning system capable of generating and automating new test cases. This system is composed of several components and steps. The first step is data collection—gathering metrics, analytics, or any available information that reflects how users interact with the application. Next, a large language model (LLM) processes this data to identify patterns and propose test cases in a predefined format. Once these test cases are generated, they undergo evaluation to check whether similar cases already exist. If they are new, they are selected and added to the regression suite.

The self-learning system would be periodically retrained with new data at the required intervals, ensuring that the regression suite is continuously updated and aligned with real user behavior.

I truly believe this topic will be of great interest, as it demonstrates a real-world example of how AI can support QA. We all know this is a hot topic right now, often raising debates about whether AI will replace QA. The key message here is that AI is not about replacement—it’s about shaping the future of QA by enabling us to do things differently, and ultimately, in a better way.