AI Data Annotation Specialist in Brazil

Rex.zone; Brasil; Teletrabalho; BRL 36 000 - 60 000; Tempo integral

Resumo da oferta

A leading tech firm is seeking a remote, full-time AI/ML data operations professional to support their LLM training pipelines. The ideal candidate will have mid-senior experience in data annotation and QA, with strong judgment skills in RLHF evaluations. Responsibilities include performing data labeling tasks, improving training data quality, and collaborating on error analysis. This role offers the opportunity to work with cutting-edge AI technology from Brazil.

Qualificações

• Mid-Senior experience in annotation, LLM evaluation, or data labeling QA.
• Strong judgment for RLHF-style comparisons and preference ranking.
• Proven ability to follow detailed annotation guidelines with high consistency.
• Clear written communication for escalations and audit notes.

Responsabilidades

• Perform AI data annotation and data labeling for LLM training datasets.
• Complete RLHF preference ranking, prompt evaluation, and QA evaluation tasks using consistent rubrics.
• Improve training data quality by reducing label noise and documenting edge cases.
• Collaborate with engineering/ops on error analysis and rubric updates.
• Support computer vision annotation (bounding boxes, polygons, segmentation) when needed.
• Handle content safety labeling aligned to policy/risk taxonomies.
• Collaborate with engineering/ops on error analysis and rubric updates.

Conhecimentos

Data annotation

Data labeling

Quality assurance (QA)

Natural Language Processing (NLP)

Computer vision annotation

Descrição da oferta de emprego

Overview

Remote, full-time AI/ML data operations role supporting LLM training pipelines from Brazil. You will deliver high-quality annotations, evaluations, and QA using clear rubrics and strict guideline compliance.

What You Will Do

• Perform AI data annotation and data labeling for LLM training datasets
• Complete RLHF preference ranking, prompt evaluation, and QA evaluation tasks using consistent rubrics
• Improve training data quality by reducing label noise and documenting edge cases
• Execute NLP tasks such as named entity recognition and classification
• Support computer vision annotation (bounding boxes, polygons, segmentation) when needed
• Handle content safety labeling aligned to policy/risk taxonomies
• Collaborate with engineering/ops on error analysis and rubric updates

Required Qualifications

• Mid-Senior experience in annotation, LLM evaluation, or data labeling QA
• Strong judgment for RLHF-style comparisons and preference ranking
• Proven ability to follow detailed annotation guidelines with high consistency
• Clear written communication for escalations and audit notes

How To Apply

Apply on Rex.zone and complete any screening tasks (prompt evaluation, RLHF ranking, or QA samples).