Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
TL;DR
Data Annotation Specialist (AI): Evaluating, stress-testing, and improving Large Language Models through structured feedback and data labeling with an accent on analytical judgment and quality assurance. Focus on identifying model failure modes, creating high-quality training datasets, and maintaining consistency across complex multimodal tasks.
Location: Must be based in Canada
Salary: CAD $30/hour
Company
A leading security-first enterprise AI company building cutting-edge foundation models and end-to-end products to solve real-world business problems.
What you will do
- Evaluate and rank model outputs based on accuracy, helpfulness, and safety.
- Stress-test models to surface failure modes and unsafe behaviors.
- Author high-quality prompts and responses to build training datasets.
- Design and apply grading rubrics to ensure consistent annotation quality.
- Annotate and rectify inaccuracies across text, image, and structured data formats.
- Communicate performance trends and quality issues to cross-functional teams.
Requirements
- Must be based in Canada
- 1+ years of experience in AI data annotation, LLM evaluation, or related analytical roles.
- Strong command of written English with the ability to provide clear justifications for evaluations.
- Comfort working with structured data formats including JSON, CSV, Markdown, XML, and YAML.
- Ability to commit to 16 hours per week.
- Must provide own laptop (BYOD).
Culture & Benefits
- Opportunity to contribute to frontier AI technology development.
- Flexible, remote-first work environment.
- Exposure to experimental and evolving AI task types.
- Collaborative work within a global, asynchronous team.
Hiring process
- Resume and writing sample review.
- Virtual annotation test covering language-based tasks.
- Video screen with the Operations Team.
Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →