PDF Structuring and Annotation Specialist

About OpenTrain

OpenTrain AI is the hiring and contracting organization for this project. OpenTrain is the #1 platform for finding and building careers in AI training and data labeling, helping contributors discover projects, build their profiles, and apply in minutes.

• Free OpenTrain account
• Remote contract work
• Part-time opportunity with less than 20 hours per week

About AI Training and Document Annotation

AI training is the human side of building artificial intelligence. People prepare and review examples that help AI systems understand text, images, documents, and structured information.

In this project, you will work with PDF content and create precise annotations that represent both document layout and text. Your work will help produce clean, structured data for AI systems.

• Contribute to cutting-edge AI development
• Use document analysis, transcription, and structured data skills
• Work remotely with a flexible part-time schedule

The Role

OpenTrain AI is seeking detail-oriented PDF Structuring and Annotation contractors with C1 or C2 English proficiency and a basic technical background in HTML or JSON formatting. The role focuses on identifying document elements, accurately representing their hierarchy, and converting PDF content into structured data.

This is an entry-level, part-time contractor opportunity requiring less than 20 hours per week. The project is available to candidates located in India or the Philippines.

• Employment type: Contractor and part-time
• Pay: $7 USD per hour
• Location eligibility: India and the Philippines
• Language: English
• Experience level: Entry level

What You’ll Do

You will annotate PDF documents by identifying and labeling their visible structure and content. Accuracy is essential because each document must be represented consistently in a clean, structured format.

• Draw bounding boxes around headings, paragraphs, figures, tables, images, charts, and other document elements
• Categorize elements such as section headers, text blocks, figures, charts, and images
• Assign each element to the correct document hierarchy, including sections and subsections
• Transcribe all visible English text exactly as it appears
• Describe images and figures
• Convert tables into well-formatted JSON arrays
• Create structured representations of PDF content and layout

Requirements and Qualifications

You should be comfortable working carefully with document layouts, hierarchy, and structured information. C1 or C2 English proficiency is required, and you will need to provide an official document certifying your English level after applying and before starting the project.

• C1 or C2 English proficiency
• Experience with document annotation or transcription
• Familiarity with sections, subsections, headers, and other document structures
• Ability to draw accurate bounding boxes and label document elements
• Basic knowledge of HTML or JSON formatting
• Ability to transcribe text exactly as shown
• Skill in identifying and describing figures, charts, and images
• Experience converting tables into structured JSON arrays

Interview Assessment

Before completing the live chat interview, applicants will be asked to submit a JSON response to a table annotation test. The test evaluates whether you can follow a required structure and accurately represent tabular content.

• Convert a sample table into structured JSON
• Include rowCount and columnCount fields
• Provide a cells array containing cell objects
• Identify headers with kind set to columnHeader and data cells with kind set to bodyCell
• Include rowIndex, columnIndex, and content for each cell
• Use row indexing beginning at 0 for the header row
• Submit clean JSON that follows the specified format exactly

How to Apply

Create a free OpenTrain account and apply through the platform. Be prepared to demonstrate your JSON formatting ability during the interview process and provide official C1 or C2 English certification after applying and before the project begins.

• Apply through OpenTrain AI
• Complete the required table annotation test before the live chat interview
• Provide English proficiency documentation if selected to continue
• Work remotely for less than 20 hours per week