English (U.S. Native) AI Trainer & Evaluator (Remote, Hourly Contractor)

CNTXT AI; New York (NY); Remote; USD 28,000 - 55,000; Part time

Job summary

CNTXT AI is seeking a fully remote, hourly contractor to support AI data and language projects with flexible hours. You will perform content generation, data annotation, LLM evaluation, and localization QA for diverse topics across time zones.

Ideal candidates are native US English speakers with exceptional editorial judgment and extreme attention to detail. Prior experience with QA workflows and AI data tasks is preferred, and reliability across time zones is essential.

Qualifications

• Native US English speaker with strong editorial judgment.
• Excellent punctuation, tone, and inclusive language.
• Ability to identify drift, ambiguity, and locale inconsistencies.
• Ability to fact-check references, dates, and names using reliable sources.

Responsibilities

• Content generation: prompts and model responses, or voice samples.
• Data annotation: label and structure documents for training data.
• LLM evaluation: assess accuracy, reasoning, coherence, and culture.
• Localization QA: ensure locale-specific accuracy across outputs.

Skills

Editorial judgment

Attention to detail

Quality assurance

Localization QA

Tools

MQM/LQA

Job description

This is a fully remote, hourly contractor role supporting AI data and language projects on a project-based, flexible hour basis. Project scopes vary and may include:

• Content generation: writing high-quality prompts and model responses, or recording high-quality voice samples, to guide AI learning across diverse topics.
• Data annotation: labeling, classifying, and structuring documents, tables, and other content to support AI training datasets.
• LLM evaluation: reviewing AI-generated responses for accuracy, reasoning quality, coherence, and cultural/linguistic appropriateness.
• Localization QA: ensuring terminology, tone, cultural nuance, and locale-specific details (units, references, names, dates) are consistently accurate across outputs.

Profile Requirements

• Native speakers of American English born and raised in the United States.
• Excellent editorial judgment in register, tone, punctuation, inclusivity, and cultural nuance, with extreme attention to detail.
• Ability to identify meaning drift, ambiguity, locale inconsistencies, and subtle errors, and to explain corrections clearly in writing.
• Ability to rigorously fact-check localized content (units, references, names, dates) using reliable sources and consistent reasoning.
• Ability to identify reasoning gaps, methodological errors, and unclear explanations even when language is fluent.
• Reliable, self-directed, and able to deliver consistent quality with clear communication and responsiveness across time zones.

Preferred Experience

• Familiarity with MQM/LQA concepts (severity, category, and root-cause thinking) for consistent quality decisions.
• Familiarity with QA workflows.
• Previous experience with AI data training, annotation, or evaluation.

About CNTXT AI

CNTXT AI builds artificial intelligence products and data solutions with a focus on making AI accurate, safe, and globally relevant for impact. Our work spans data services, custom AI solutions, and proprietary AI products, with deep expertise in Arabic-native and secure, sovereign solutions.