Remote Generalist (Brazil)

Remote Generalist (Brazil) — Full-Time (Remote)

Rex.zone is hiring Brazil-based professionals to support AI/ML data operations that improve large language model evaluation and training data quality across multiple projects.

About The Role

You will work across data labeling, RLHF preference ranking, prompt evaluation, and QA evaluation. Work may span NLP annotation (including named entity recognition), occasional computer vision annotation, and content safety labeling while following detailed rubrics and escalation rules.

What You Will Do

• Perform RLHF ranking and side-by-side comparisons of model outputs
• Evaluate prompts and model responses for helpfulness, harmlessness, factuality, and rubric adherence
• Complete QA evaluation by auditing samples and documenting edge cases
• Label text/image data (NLP + computer vision tasks as applicable)
• Escalate policy/safety concerns and contribute feedback to improve guidelines
  
  

Required Qualifications

• Mid-Senior experience in operations, analytics, QA, trust & safety, or data work
• Strong written communication and careful reading of detailed instructions
• Ability to apply rubrics consistently with high accuracy and throughput
• Comfort with ambiguity and iterative guideline updates
• Reliable remote work habits for full-time delivery
  
  

Preferred Qualifications

• Experience with AI data labeling platforms and QA dashboards
• Familiarity with RLHF, prompt evaluation, and LLM evaluation
• NLP annotation experience (e.g., named entity recognition, taxonomy/intent labeling)
• Content moderation/content safety labeling experience
• Quality processes (e.g., sampling plans, inter-annotator agreement, escalation protocols)
  
  

How Success Is Measured

Guideline adherence, QA pass rates, consistency with rubrics, low error rates, and clear documentation that reduces label noise and improves model performance.

How To Apply

Visit Rex.zone to explore current requirements and apply.