Personal Account AI Response Rater (German, Japanese, Korean, French)

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We help freelancers grow lasting AI-training portfolios, discover projects that match their skills, and manage work across contracts in one place.

OpenTrain is the hiring and contracting organization for this role. We focus on remote, flexible opportunities that let people get paid for the human work behind modern AI systems.

About AI training work

AI training (also called data labeling or human feedback work) is the human side of building intelligent assistants. Contributors annotate, rate, and critique model outputs so systems learn to produce accurate, helpful, and appropriately personalized responses.

This role centers on language-aware judgment and clear written feedback — core tasks in RLHF-style projects that shape how assistants perform for real users.

The role

You will work as a Personal Account AI Response Rater, evaluating AI-generated replies against a professional user’s email history and account activity. Your judgments will help the model learn to respond with correct facts, relevant recommendations, and appropriate personalization for business contexts.

This is a part-time contractor role requiring 20+ hours per week and available to candidates based in the United States only.

What you'll do

You will compare AI responses, assess personalization quality, and provide detailed feedback following project guidelines. Work is text-focused and uses RLHF-style evaluation and rating tasks.

• Evaluate AI-generated responses using professional email history and Google Workspace account activity.
• Assess relevance, accuracy, helpfulness, and whether responses are properly personalized to the user.
• Identify incorrect personalization, unsupported assumptions, and irrelevant or unsafe recommendations.
• Compare multiple outputs and choose which provides the better user experience.
• Write clear, detailed explanations for your evaluation decisions and follow confidentiality rules.

Requirements

Candidates must meet all technical and language prerequisites before applying. These are essential because the work requires reading and judging real business email context in the target language.

• Native-level proficiency in one target language: German, Japanese, Korean, or French.
• A well-established Google Workspace business inbox used for professional communication (Gmail).
• Gmail display language set to the target project language and most inbox emails in that language.
• Strong analytical thinking, attention to detail, and excellent written communication skills.
• Comfort reviewing business email history and identifying unsupported assumptions.
• Ability to work independently as a contractor for 20+ hours per week while following data privacy guidelines.

Helpful background & skills

The following are preferred but not strictly required. They help you hit the ground running and produce higher-quality feedback.

• Experience in AI evaluation, data annotation, content review, or QA.
• Prior RLHF or response-rating experience and familiarity with evaluation tasks.
• A bachelor's degree or equivalent practical experience.