AI Trainer
I contributed to a computer vision project focused on enhancing AI image recognition models through high-precision data preparation. My primary responsibility involved the meticulous application of Bounding Box annotations to identify and categorize specific objects within complex visual environments. The scope of my work included executing dual-role tasks where I authored complex prompts from a user perspective and crafted accurate, high-quality responses from the assistant's perspective to train conversational AI. I adhered to strict quality measures, cross-verifying both visual labels and text outputs for accuracy and adherence to project rubrics.