Testing with GenAI, but Make It Understandable!
Generative AI is transforming software testing, introducing powerful new capabilities while also creating complexity that can be overwhelming. This session presents the key concepts from the ISTQB Testing with Generative AI Syllabus in a clear and practical way. Participants will gain an understanding of LLMs, learn how to craft and refine prompts to support testing tasks, and receive an overview of risks such as hallucinations, bias, and data privacy issues, along with strategies for responsible and compliant use. The talk also covers technical aspects, including infrastructure, fine-tuning, and operational considerations, as well as organizational approaches for integrating GenAI efficiently into software testing processes.
Value for the audience:
Participants will walk away with:
- A clear, practical understanding of how Generative AI fits into software testing, based on the structured ISTQB CT-GenAI syllabus. Complex topics like LLMs, prompt engineering, and fine-tuning will be explained in a way that testers, test managers, and quality leaders can directly relate to their work.
- Attendees will learn techniques to craft and refine prompts for testing tasks.
- The session highlights key risks (hallucinations, bias, non-determinism, and data privacy) and provides concrete strategies for responsible and compliant use of GenAI in testing.
- Beyond the basics, the talk covers infrastructure and LLMOps considerations, as well as change management and adoption strategies. This ensures participants not only understand the technology, but also how to integrate it sustainably within their QA organizations.
Problems addressed:
Complexity and Uncertainty: Testers often find GenAI concepts (LLMs, fine-tuning, multimodality, RAG) overwhelming and disconnected from their day-to-day testing activities.
Risk Blindness: Many teams jump into using GenAI without understanding risks such as hallucinations, bias, data privacy and security concerns.
Integration Gap: Organizations struggle to move beyond experiments with AI toward sustainable integration in infrastructure, processes, and skills development for QA teams.
Talk language: English
Level: Newcomer
Target group: testers, test managers, test automation engineers, test analysts and quality leaders
Company:
mgm technology partners gmbh
Diplom Informatik Lilia Gargouri