What is it about?

AGORA is a new AI tool that automatically creates software test cases. In our study, industry professionals found that AGORA produces tests comparable to human-written ones but 90% faster. This breakthrough helps companies develop better software more quickly while maintaining quality.

Featured Image

Why is it important?

AGORA is the first solution to fully automate acceptance test creation from use cases using Large Language Models. Unlike previous approaches that only automated unit testing or required significant human input, AGORA bridges requirements and test cases with minimal effort. Its timing is perfect for today's agile development environments where testing speed is crucial. By reducing test creation time by 90% while maintaining quality, AGORA removes a major development bottleneck. What sets AGORA apart is its ability to truly understand use case scenarios and generate human-readable test cases that teams can easily review and maintain. With validation from industry professionals, it offers an immediately applicable solution for companies facing pressure to deliver quality software faster.

Perspectives

Developing AGORA has been deeply rewarding as it addresses a pain point I've observed throughout my career in software engineering. During our industry evaluations, seeing professionals' excitement when they realized they could generate in minutes what typically took hours was genuinely moving. One tester exclaimed they had "waited their entire career for this." This research has connected me with testing communities I wouldn't otherwise have engaged with, leading to valuable conversations about the future of quality assurance. I hope this work not only makes testing more efficient but also helps elevate the role of QA professionals by freeing them to focus on more creative aspects of ensuring software quality.

Dr Gabriele De Vito
Universita degli Studi di Salerno

Read the Original

This page is a summary of: AGORA: An Approach for Generating Acceptance Test Cases from Use Cases, August 2024, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/seaa64295.2024.00027.
You can read the full text:

Read

Contributors

The following have contributed to this page