Large-Scale Image & Text Data Annotation for Machine Learning Models
Worked on a multi-domain data labeling and annotation project supporting the training and evaluation of machine learning and large language models. The project involved annotating both image and text datasets at scale while maintaining strict quality standards. Key responsibilities included: Annotating images using bounding boxes, polygons, and segmentation masks for object detection and classification tasks Performing Named Entity Recognition (NER), text classification, question-answer pair validation, and summarization labeling Reviewing and validating peer annotations to ensure consistency and accuracy Following detailed annotation guidelines and resolving edge cases through feedback loops Handling datasets ranging from 10,000+ data points per task Quality was maintained through multi-pass reviews, gold-standard comparisons, and adherence to platform-specific QA benchmarks (precision, recall, and inter-annotator agreement).