Multimodal AI Document Understanding – Data Annotation
I built multimodal document understanding systems by labeling scanned document data for classification and automated extraction. Tasks included annotating and validating both text and image content for the development of OCR and vision-based models. I combined manual annotation with automated pipeline approaches for efficient processing. • Labeled and validated documents for OCR model training. • Annotated multi-format content for text-image understanding models. • Contributed to dataset curation for document classification and extraction. • Collaborated with team members to verify annotation consistency and accuracy.