AI Training Data Annotator – Outlier & Alignerr
Performed AI training and data evaluation tasks across multiple project types on the Outlier and Alignerr platforms, including image quality assessment, sketch-to-photorealistic conversion evaluation, bounding box annotation accuracy review, and comparative response ranking for language model outputs. Applied structured reasoning to identify factual inaccuracies, visual inconsistencies, and annotation errors, providing detailed written justifications for each verdict. Worked with both visual and text-based data modalities across tasks in English and Portuguese.