AI Personalization Evaluator (German)

You will evaluate how a conversational AI model uses information from earlier conversations and connected activity to make its replies more relevant and helpful. The work combines German-language review, prompt writing, careful comparison, and evidence-based judgment.

You will also check supporting information, write detailed annotations, and remove evaluation conversations after review to maintain good data hygiene.

• Create and run multi-turn prompts, usually covering one to five turns, using personal experiences and context.
• Check whether model responses use personal information appropriately and identify unsupported claims, poor inferences, hallucinations, or forced connections.
• Rate grounding, integration, helpfulness, ease of use, enjoyment, naturalness, and overexplaining.
• Compare two model responses side by side and explain your decision using specific conversation turns.
• Extract and verify supporting information to confirm that summaries and relevant data sources were used correctly.
• Provide constructive feedback and detailed annotations.
• Delete evaluation conversations after review.

This is a remote, part-time contractor engagement expected to last three months. Candidates worldwide can apply.

• $15 per hour.
• 20+ hours per week, with availability of at least four hours per day and up to 40 hours per week.
• Four hours of working-time overlap with Pacific Time is required.
• Fluent German reading and writing ability.
• Intermediate experience level with strong analytical judgment for nuanced or unclear AI responses.
• Experience creating creative prompts that test model behavior using personal context.
• Ability to identify incorrect personalization, unsupported inferences, and unnatural use of personal data.
• Clear, concise, structured writing with precise, evidence-based reasoning.
• Willingness to use a primary personal account and enable relevant personal data sources for evaluation.
• A desktop or laptop, reliable internet connection, and ability to work independently.
• A degree or equivalent experience in policy, law, ethics, linguistics, journalism, computer science, or a related analytical field is helpful.
• Experience in data annotation, AI quality evaluation, content moderation, or similar review work is valuable.

You apply on OpenTrain with your resume and then complete the application on the hiring site.

AI training is the human work behind systems that learn from examples, including writing prompts, reviewing model responses, and rating whether answers are accurate and useful. People with strong language skills and careful judgment help shape how these systems respond.