Entry Level Data Labeling Specialist (United States, Remote)

**About Rex.zone** Rex.zone connects remote contributors with AI data operations work, including data labeling, RLHF evaluation, prompt evaluation, and quality assurance for NLP, computer vision, and large language model training pipelines. **Role Summary** You will label and evaluate training data that improves AI/ML systems, focusing on consistent guideline adherence and measurable training data quality. Work includes text, images, and multimodal evaluation tasks, plus structured QA workflows. **Key Responsibilities** * Perform data labeling and data annotation across NLP, computer vision, and multimodal datasets * Apply annotation guidelines and document edge cases for adjudication * Execute QA evaluation workflows (spot checks, consensus review, error categorization) * Support RLHF by rating responses, ranking candidates, and flagging hallucinations or policy risks * Conduct prompt evaluation and evaluation set curation to improve model performance * Maintain throughput and accuracy targets while protecting sensitive data **What You Will Work On** * LLM evaluation: helpfulness, harmlessness, honesty, instruction following * Content safety labeling: policy\-based classification, severity rating, escalation signals * NLP labeling: named entity recognition (NER), topic tagging, intent detection, factuality checks * Computer vision annotation: bounding boxes, polygons, segmentation masks, image QA **Qualifications** * Comfort with detailed rules, annotation rubrics, and high\-precision work * Strong written communication for documenting rationales and disagreements * Ability to learn new labeling tools and taxonomy updates quickly * Familiarity with annotation concepts (NER, CV annotation, RLHF) is helpful but not required **How To Apply** Explore current entry level data labeling jobs in the United States on Rex.zone and submit your application with availability and any relevant annotation, QA, or evaluation experience.