Remote Data Annotator

Rex.zone is hiring for remote data annotator work supporting modern AI/ML training pipelines. You will create high-signal labeled datasets used in NLP, computer vision, and content safety systems, and contribute directly to model reliability through consistent guideline application and QA evaluation.

• Annotate text for NLP tasks (classification, named entity recognition, summarization evaluation)
• Perform prompt evaluation and LLM response scoring using rubrics
• Complete RLHF preference ranking and comparison tasks
• Label images for computer vision annotation (bounding boxes, polygons, segmentation)
• Document edge cases and provide clear notes to improve annotation guidelines
• Run QA checks, support audits, and participate in calibration sessions

• Guidelines compliance and accuracy on gold tasks
• Inter-annotator agreement and audit outcomes
• Low rework rates and clear edge-case documentation

• Experience with structured labeling workflows and careful, consistent decision‑making
• Comfort working in web‑based annotation tools and tracking tasks
• Strong written communication for remote collaboration and escalation of ambiguous cases

• Experience with RLHF, content safety labeling, or policy enforcement labeling
• Familiarity with NLP evaluation tasks (NER, instruction‑following, retrieval relevance)
• Exposure to computer vision annotation and dataset QA concepts (gold sets, sampling, agreement)

This is a role that is explicitly Remote and can be performed from Portland, Oregon using online tools and remote team communication.