Project Hedgehog
On Project Hedgehog, I contributed to AI training and data annotation across multiple workflows focused on improving model accuracy and output quality. My responsibilities included image labeling (such as brain image tagging), annotating entities like companies and locations, and completing generalist tasks across diverse datasets. I also performed human-to-human (H2H) evaluations to assess response quality and consistency. A key part of my role involved identifying low-quality or incorrect outputs and completing isolation, removal, and replacement tasks to refine training data. Additionally, I worked on audio transcription, ensuring clarity and accuracy in converting speech to text. This project required strong attention to detail, adherence to evolving guidelines, and the ability to maintain high-quality standards across high-volume tasks.