This research addresses a critical gap in software testing: creating meaningful test cases that align with business needs, not just code coverage.

Key Findings:

• Industry survey revealed the primary challenge is aligning tests with business requirements, not writing test scripts
• Researchers created a specialized dataset mapping business use cases to high-level test cases
• LLMs were trained/fine-tuned to automatically generate high-level test cases from business requirements

This innovation streamlines the testing process by automating a cognitively demanding task, potentially reducing testing costs while improving quality and security coverage in software development lifecycles.

Automatic High-Level Test Case Generation using Large Language Models