Senior Data Annotation Specialist at Alegion Inc. (video temporal tracking and QA).
Produced frame-by-frame temporal object tracking annotations for video datasets used in production ML training. Ensured inter-annotator agreement and consistent object persistence across multi-person projects. Applied QA review processes to maintain annotation integrity and accuracy for downstream model performance. • Labeled 200+ video clips weekly using SuperAnnotate and Roboflow. • Maintained accuracy exceeding 98.5% based on internal QA audits. • Managed inter-annotator agreement for consistent tracking across frames. • Refined guidelines based on ML engineer feedback to reduce ambiguity by 30%.