Korean PDF Annotator and Transcriptionist

Help build high-quality Korean-language training data for document AI by creating detailed structural maps and exact transcriptions of real, publicly available PDF pages. The work focuses on complex materials that are difficult for automated systems to interpret, including handwriting, dense multi-column layouts, tables, diagrams, and mixed-script pages from newspapers, textbooks, examinations, flyers, forms, manuals, menus, brochures, notices, and worksheets. All annotation and transcription work is performed in Korean, and every deliverable is created by human experts.

• Find publicly accessible Korean PDFs within assigned document types that include at least one multimodal element, such as an image, table, diagram, or handwriting, and record the source.
• Identify, bound, and classify every meaningful page region, including titles, headings, paragraphs, lists, tables, figures, diagrams, captions, formulas, questions, and answer fields.
• Assign reading-order indexes and link related regions to their parent figures or tables using component identifiers.
• Transcribe all text exactly as shown in Hangul, including hanja and handwritten content, and flag text that cannot be read from the source.
• Capture page metadata, including language, document type, source, page dimensions, and indicators for tables, formulas, and handwriting.
• Review a colleague''s work end to end as an experienced annotator.

• Native Korean fluency and full command of Hangul are required. Korean is the required working language for all annotation and transcription.
• Ability to work with hanja in older or formal documents.
• Experience in bilingual transcription, translation, editorial work, or AI training data is preferred, ideally including reviewer experience.
• Exceptional character-level accuracy, including careful handling of individual Hangul jamo.
• A systematic approach to applying a consistent document taxonomy across many pages.
• Comfort working with unfamiliar layouts, including multi-column newspapers, exam papers, and handwritten forms.

• Annotation, labeling, grading, or bilingual evaluation for AI training datasets.
• MTPE, subtitling, bilingual QA, OCR correction, or post-editing.
• Korean-language typesetting, copy-editing, proofreading, or digitization.
• Unicode normalization, Korean input methods, Hangul jamo composition, and hanja handling.

• Capture, correctly bound, and correctly classify every meaningful region on each page.
• Reflect the actual reading order, including across columns.
• Produce character-for-character transcriptions in Hangul rather than romanization.
• Deliver work that passes second-expert review on the first submission.
• Contribute documents that expand layout diversity rather than repeating existing templates.

• Remote, hourly engagement.
• Open only to contributors based in South Korea, Japan, India, the United States, Canada, or Western Europe.
• Accuracy is the primary performance measure, with handling time also tracked.

$33.58 per hour.