Programme launch offer - Save 20% Book Now

Track Talk W3

DocuMentor: Enhancing Test Documentation with AI Feedback

Richard Bishop

Liam Patience

10:30 - 11:15 CEST Wednesday 4th June

In March 2024, our team participated in a Google hackathon with the objective of exploring how Generative AI (GenAI) could be used to assist engineers. We set ourselves the challenge of using GenAI to enhance quality assurance processes by using a Large Language Model (LLM) to assess and improve testing documents.

We developed an innovative application, nicknamed “DocuMentor” to compare test plans against predefined criteria for a “perfect” test plan, providing valuable feedback to users. The application allows users to upload a document, which is then compared to the “golden standard” and feedback is generated to help improve the quality of the test plan.

Our solution used Google’s Gemini Pro LLM. However, upon reflection, we realised that using a smaller, more specialised model might have been more appropriate for our specific needs. Since the hackathon we’ve improved performance and document quality by using smaller, faster LLMs. The hackathon and subsequent development of the application, provided us with insights into the application of GenAI as an assistant.

One of the key takeaways from our experience was the realisation that by creating other “golden standards,” we can check almost any type of document against a defined standard. This flexibility opens up numerous possibilities for improving document quality across various domains. After the hackathon, we added collaborative chat functionality to our application, enhancing its effectiveness and helping engineers to improve overall document quality.

Attendees of our talk will benefit from our insights into the practical application of GenAI in quality assurance, they’ll hear challenges we faced and the solutions we developed.

We are looking forward to demonstrate our application and share our experience with EuroSTAR Conference delegates encouraging discussion about how AI can transform quality assurance.