From Guesswork to Greatness: Building Better AI Products Through Smart Testing
Testing AI products requires a unique approach, and in this session, Maaria Tiensivu will share her proven framework that goes beyond traditional testing methods. You'll learn her signature three-persona testing strategy that includes ideal users, unlikely users, and those who can find edge cases – giving you a comprehensive view of your AI product's performance.
What makes Maaria's session particularly valuable is its focus on AI-specific testing challenges. You'll learn how to evaluate patterns rather than individual cases, understand the relationship between inputs and variable outputs, and create testing plans that account for AI's unique characteristics. She breaks down testing into clear, actionable components: defining your "why," identifying "who" to test with, determining "how" to test, and planning what to do with the results.
Whether you're launching a new AI product or improving an existing one, you'll walk away with concrete tools to create effective testing plans, select appropriate testers, and evaluate both quantitative and qualitative feedback. This practical, hands-on approach will help you build more reliable AI products that truly serve your users' needs