PDF Annotation & Transcription Experts – Telugu

Mercor is creating training data that helps document‑AI systems understand languages and scripts that are rarely seen by current models. The position of PDF Annotation & Transcription Expert – Telugu focuses on building a high‑quality corpus of real Telugu documents, ranging from newspapers and textbooks to flyers and forms, so that future AI can parse complex layouts and handwritten text. Each day you will locate a publicly available Telugu PDF that contains at least one multimodal element such as an image, table, diagram or handwriting, and note its source. You will then mark every meaningful region on the page—titles, headings, paragraphs, lists, tables, figures, captions, formulas, questions and answer fields—assigning a component type, a reading‑order index and any parent‑child relationships. The text in each region must be reproduced exactly as it appears in the original script, with illegible areas flagged, and page metadata recorded. After completing a task, you will review a peer’s work end‑to‑end, and experienced annotators will also perform reviews. The role requires native‑level fluency in Telugu, complete mastery of the script’s diacritics and conjunct forms, and prior experience with document‑related work such as annotation, transcription, translation, localization, subtitling, proofreading, journalism or regional‑language data review. Candidates must be detail‑oriented, able to maintain character‑level accuracy, apply a consistent taxonomy across many pages, and feel comfortable with unfamiliar layouts like multi‑column newspapers or handwritten forms. Additional expertise that is advantageous includes work on regional‑language AI data, MTPE, OCR correction, typesetting, copy‑editing, Unicode normalization or Telugu input methods. The project emphasizes accuracy above speed, with handling time monitored, and success is measured by fully captured, correctly bounded and typed regions, precise reading order, and transcriptions that match the source character for character on the first review.