Remote Data Annotator

About The Role

Rex.zone is seeking experienced Mid-Senior Remote Data Annotators to generate and verify top-quality training datasets essential for the development and assessment of AI and machine learning models. This role involves working with textual, visual, and multimodal data, applying structured rubric-based decisions along with thorough documentation to facilitate measurable enhancements in model performance.

Responsibilities

• Label and verify datasets for natural language processing and computer vision tasks, including named entity recognition, classification, and image bounding box or segmentation annotations.
• Conduct Reinforcement Learning from Human Feedback (RLHF) assessments and prompt evaluations by scoring and ranking model-generated outputs according to established guidelines.
• Perform quality assurance reviews such as audit sampling, calibration, and resolving annotation disagreements, ensuring compliance with annotation standards.
• Assist with content safety classifications leveraging policy-based categories and safety evaluation rubrics.
• Work collaboratively in asynchronous settings with project leads to fulfill requirements in throughput, accuracy, and auditability within large language model training operations.

Requirements

• Demonstrated mid to senior level expertise in data annotation, data labeling, quality assurance evaluations, or involvement in large language model evaluation workflows.
• Exceptional attention to detail and the capacity to consistently address ambiguous or edge cases with logical reasoning.
• Strong written communication skills to document unique cases and reasoning behind annotations.
• Proficiency in using annotation software and spreadsheet-like review interfaces.

How To Apply

Interested applicants should visit Rex.zone and search for "Remote Data Annotator Jobs in Berlin" to access available assignments and submit an application.