AI Output Evaluator & User Tester (ReadMe App)
I worked on evaluating and improving AI-generated book recommendations in a personality-based reading application designed for children. My role involved reviewing text-based outputs, rating their relevance and accuracy, and ensuring recommendations aligned with user profiles and behavioral patterns. I applied structured evaluation guidelines to assess output quality, identify inconsistencies, and refine recommendation logic through iterative feedback. In addition, I conducted user testing to analyze how recommendations performed in real-world scenarios, using insights to improve both model performance and overall user experience. • Evaluated AI-generated recommendations for relevance, accuracy, and personalization quality • Rated outputs using structured guidelines and consistency checks • Identified mismatches between user personality data and recommended content • Conducted user testing to analyze behavior and recommendation effectiveness • Provided iterative feedback to improve recommendation logic and system performance • Supported testing workflows using Firebase (data handling) and Figma (user flow validation)