Entry Level Data Labeling Specialist - Remote

About This Role

This entry-level role involves labeling and evaluating training data to enhance AI/ML systems, with a strong focus on consistent guideline adherence and data quality. The specialist will work with text, images, and multimodal evaluation tasks, as well as structured QA workflows.

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

Requirements

• 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

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

Application Process

• Submit your application with availability and any relevant annotation, QA, or evaluation experience

About the Company

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.