RAG Context Quality Specialist

Globik AI; United States; À distance; USD 17 000 - 25 000; Plein temps

Résumé du poste

Globik AI is seeking a RAG Context Quality Specialist to evaluate whether the context supplied to a language model is clear, useful, complete, and appropriate for answer generation.

You will inspect retrieved context and determine whether it provides the right information in a usable form, identifying noise, redundancy, missing context, and conflicting information. This role requires careful analysis and clear documentation of findings.

Qualifications

• Experience with RAG search information retrieval, research content QA or AI evaluation is a plus.

Responsabilités

• Review context supplied to AI models before answer generation.
• Assess whether the context contains useful information for the query.
• Identify irrelevant or noisy passages.
• Identify missing information and insufficient context.
• Flag duplicated or conflicting source content.
• Evaluate whether context is organized and understandable when required.
• Categorize context-quality issues.
• Document difficult cases and provide structured feedback.
• Maintain consistent evaluation quality across large datasets.

Connaissances

RAG evaluation

Description du poste

Globik AI is seeking RAG Context Quality Specialists to evaluate whether the context supplied to a language model is clear, useful, complete, and appropriate for answer generation.

You will inspect retrieved context and determine whether it provides the right information in a usable form, identifying noise, redundancy, missing context, and conflicting information.

Responsibilities

• Review context supplied to AI models before answer generation.
• Assess whether the context contains useful information for the query.
• Identify irrelevant or noisy passages.
• Identify missing information and insufficient context.
• Flag duplicated or conflicting source content.
• Evaluate whether context is organized and understandable when required.
• Categorize context-quality issues.
• Document difficult cases and provide structured feedback.
• Maintain consistent evaluation quality across large datasets.

Qualifications

• Experience with RAG search information retrieval research content QA or AI evaluation is a plus.

• Follow project-specific RAG evaluation rubrics precisely.
• Clearly document difficult or ambiguous cases.
• Participate in calibration exercises when required.
• Regular quality reviews may be conducted throughout the project.
• Selected candidates may be required to complete a practical RAG assessment before onboarding.
• A reliable internet connection and computer are required.

Pay

$12–$18/task

Apply by 2026-12-15.