Male Fertility Data Annotation Specialist

You will review and correct annotations used to train an AI-assisted semen analyzer. The work takes place on grayscale, computationally reconstructed microscopy videos, with a paid onboarding session to explain how sperm appear in this image format.

All annotation work is completed in Encord. Short video sequences are about two seconds long and may contain hundreds of densely packed cells.

• Review rotated bounding boxes around individual sperm in grayscale microscopy frames.
• Add missed sperm, remove false positives such as debris, round cells, and image artefacts, and correct box orientation.
• Confirm or correct motility classes using the WHO 6th edition (2021): progressive (PR), non-progressive (NP), and immotile (IM).
• Count sperm in defined regions when needed for comparison with model output.
• Write short notes about ambiguous cells, image-quality problems, and unclear annotation guidance.
• Possibly review sperm morphology, including head, midpiece, and tail, using Kruger and WHO-strict criteria in a later feasibility phase.

This is contractor work for an experienced semen-analysis professional. The role is open to candidates in the countries listed for the project, and country of residence is required for data-protection review.

• Pay: $6 to $14 USD per hour, with an $8 USD per hour listed rate.
• Experience: at least two years of hands-on manual semen analysis in a clinical andrology, IVF, or reproductive-medicine laboratory.
• Required experience includes concentration testing with a hemocytometer and motility grading under WHO 5th or 6th edition guidance.
• Written English proficiency is required for notes and questions about the guidelines.
• You must provide a CV and verifiable lab experience, including employer, role, and dates.
• You need careful, consistent habits for repetitive visual work and the ability to follow written guidelines exactly.
• You need reliable internet and a desktop or laptop with a screen of at least 1080p. Phone and tablet work is not supported.
• Before accessing project data, you must sign an NDA, IP assignment, and data-protection agreement.
• Preferred qualifications include an embryology or andrology credential, EQA or proficiency-testing participation, Kruger strict morphology experience, CASA experience, or prior use of Label Studio, Encord, CVAT, V7, or a similar annotation tool.

The data is de-identified research data. You must not try to identify any participant, download or copy images, take screenshots, use AI tools, or involve third parties with project data.

Before production work, you will complete a paid calibration task lasting about two to three hours. Your agreement with expert reference labels will be reviewed regularly, and you will receive brief guideline training and confirm it in a signed training record.

• Work only inside the project annotation platform.
• Do not download, screenshot, or copy project images.
• Do not use AI tools or third parties on project data.
• Include your weekly availability and time zone when applying.
• Include your credentials and any EQA or proficiency-testing participation.
• Describe the laboratories where you worked, your years of experience, approximate work volume, and the WHO edition you used.

Apply on OpenTrain. The employer reviews applications there.

AI training is the human work behind systems that learn from labeled examples and expert reviews. In this role, your clinical knowledge helps create accurate examples for a medical AI system, which is why specialized experience is required.