AI English Linguist - Remote Data & Evaluation Expert

Obsidian; Greater London; On-site; GBP 44,000 - 67,000; Full time

Job summary

Mercor is seeking experienced English Linguistic Experts to evaluate LLM performance on English tasks and curate high-quality linguistic data for training and improving model understanding and generation.

You will edit, refine, and annotate English content, identify errors and gaps, and provide linguistic feedback to guide ongoing AI model training while working remotely on a distributed team.

Qualifications

• Native English proficiency with strong grammar, syntax, vocabulary and style.
• Experience editing, proofreading, writing, linguistic annotation, language evaluation or NLP data work in English.
• Strong understanding of linguistic nuances, regional variations, tone and context-appropriate English usage.
• Excellent ability to identify subtle grammatical errors, awkward phrasing, ambiguity and unnatural language.
• Interest in AI and LLM-driven language technologies is preferred but not required.

Responsibilities

• Evaluate quality and accuracy of LLM-generated English text across use cases (summarization, dialogue, grammar correction, rewriting, reasoning).
• Edit, refine, or rewrite English content to meet high linguistic and stylistic standards.
• Annotate English text for grammatical, syntactic, semantic, and stylistic features.
• Identify model errors, inconsistencies, and language gaps, and provide feedback to improve LLM performance.
• Provide detailed linguistic feedback and analysis to guide ongoing model training efforts.

Skills

Native English

Editing & proofreading

Linguistic annotation

NLP data work

English language evaluation

Attention to detail

Remote work

Job description

Mercor is seeking experienced English Linguistic Experts to evaluate LLM performance on English tasks and curate high-quality linguistic data for training and improving model understanding and generation.

You will edit, refine, and annotate English content, identify errors and gaps, and provide linguistic feedback to guide ongoing AI model training while working remotely on a distributed team.