PDF Structuring & Annotation Specialist (C1/C2 English)

About OpenTrain

OpenTrain is the #1 platform for building careers in AI training and data labeling. We connect people to remote, flexible work that helps shape how modern AI learns and behaves.

We hire contributors directly for annotation projects. This role is managed and contracted by OpenTrain AI.

About AI training and document annotation

AI training (also called data labeling or annotation) is the human work behind model behavior: humans create and refine the examples models learn from. Document annotation is a common task in the field — structuring PDFs, transcribing text, and labeling layout elements so models can understand documents.

These projects are often remote, flexible, and accessible to people with good attention to detail and domain-relevant skills such as transcription, HTML/JSON familiarity, or previous annotation experience.

The role — what you'll do

You will convert PDF pages into a clean, structured representation of content and layout. Work is remote and part-time (less than 20 hours per week) as a contractor.

• Draw bounding boxes around document elements (headers, paragraphs, figures, tables, images).
• Label each element with its type and place it in the document hierarchy (section, subsection, etc.).
• Transcribe all visible English text exactly as shown.
• Describe figures and images where required.
• Convert tables into well-formatted JSON arrays (row/column counts and cell content).
• Use basic HTML/JSON knowledge to produce structured outputs compatible with the project format.

Project details & data types

Data type: DOCUMENT. Label types: BOUNDING_BOX and TEXT_GENERATION. You will work with scanned or native PDFs and produce bounding box annotations plus exact text transcriptions and structured JSON for tables.

• Employment types: CONTRACTOR, PART_TIME.
• Time commitment: Less than 20 hours/week.
• Pay: USD 7 per hour, paid on an hourly basis.
• Worldwide applicants may apply; work is fully remote.

Requirements

We are looking for detail-oriented contributors comfortable with English at C1/C2 level and familiar with document structure and basic structured-data formats.

• C1/C2 English proficiency is required; you will need to provide an official certificate after applying and before starting.
• Experience with document annotation, transcription, or structured data creation.
• Ability to accurately draw bounding boxes and label elements by type and hierarchy.
• Ability to transcribe text exactly as it appears in the PDF.
• Familiarity with HTML or JSON formatting and converting tables into JSON arrays.
• Skill in identifying and describing figures, charts, and images.

Interview & mandatory test task

During the interview you will complete a short live chat and must submit the JSON test response below before finishing that conversation. This verifies your ability to convert tables into the required structured format.

• Test Task: Table Annotation (JSON Format) — Convert the sample table into JSON with rowCount, columnCount, and a cells array.
• Table 1: AI Adoption Rates by Industry | Industry | Adoption Rate | |--------------|----------------| | Healthcare | 76% | | Finance | 65% | | Retail | 48% |
• Required JSON fields for each cell object: kind ("columnHeader" or "bodyCell"), rowIndex (0 = header), columnIndex (0 = first column), content (text).
• Submit a clean JSON response that exactly matches the specified structure before completing the live chat interview.

Who should apply

This role is ideal for entry-level annotators, transcriptionists, editors, or anyone with an eye for layout and basic HTML/JSON experience. If you meet the language requirement and enjoy structured, detail-focused work, apply.

• Good fit: careful transcribers, document editors, data-entry specialists, content-structure enthusiasts.
• Not required: prior machine-learning experience — we provide project-specific guidelines.

How to apply

Apply through OpenTrain. Be prepared to demonstrate C1/C2 English proficiency (certificate required before starting) and to complete the JSON test during the interview. Successful applicants will be onboarded with project guidelines and examples.

• Include any previous annotation, transcription, or document-structuring work in your application.
• You will receive task-specific instructions and examples after hire; follow them exactly to ensure consistent structured outputs.