BrowserStack QA Online meetup - Krakow

Testing the Untestable: Lessons from AI Evaluation BrowserStack Tech-Leader Series | Krakow Chapter (Virtual) Traditional QA relies on a predictable rule: input $X$ equals output $Y$. Generative AI has broken that rule completely. When dealing with non-deterministic systems like Large Language Models (LLMs), outputs are fluid and probabilistic. How do you automate testing for a system designed to never give the exact same answer twice? How do you reliably catch hallucinations or systemic quality drops before they hit production? This session strips away the AI hype to look at the raw mechanics of AI evaluation—moving from rigid binary code assertions to dynamic data validation, semantic evaluation, and automated statistical oversight. 🔍 Core Focus Areas Moving Beyond Pass/Fail: Adapting traditional validation frameworks to handle unpredictable, fluid model responses. Scale-Level NLP Evaluation: Tactical lessons from auditing, refining, and grading millions of natural language processing (NLP) data points for enterprise AI systems. The New QA Toolkit: How data profiling, mathematical analysis, and statistical validation are becoming mandatory quality engineering skills. 🎙️ About the Speaker Jasna Bogunović Marković | Advanced AI Data Trainer & QA Lead at Invisible Technologies Jasna combines over two decades of advanced mathematics and statistical analysis expertise with cutting-edge AI quality validation. Based in Serbia, she leads error analysis, NLP data refinement, and pre-deployment LLM validation for global technology enterprises. She specializes in translating non-deterministic, chaotic AI behaviors into structured, measurable production engineering standards. 📅 Event Logistics Format: Online / Virtual Meetup Target Audience: Krakow Quality Engineering & QA Community Date: Tuesday, June 30th, 2026 Time: 6:00 PM (Krakow CEST)