You will evaluate how well an AI system uses personal context to make responses more relevant and helpful. You will create realistic multi-turn conversations, review model responses, compare outputs, and write structured feedback for an AI training project.
The work includes checking whether personalization is grounded in the available information, identifying unsupported inferences or hallucinations, and judging whether personal details are used naturally rather than explained awkwardly.
- Create and run multi-turn prompts based on personal experiences and context.
- Check whether responses use personal information appropriately and accurately.
- Identify unsupported inferences, hallucinations, forced connections, and other personalization problems.
- Compare and rank two model responses for helpfulness, ease of use, enjoyment, and overall quality.
- Write clear rationales that refer to specific conversation turns.
- Review debugging information and data-source use, then delete evaluation conversations to keep chat history clean.
What It Pays and Takes
This is a remote, part-time contractor engagement lasting one month. The role focuses on Polish-language evaluation and is open to remote contractors where permitted.
- Pay: $20 per hour.
- Availability: At least 4 hours per day, 20 or more hours per week, and up to 40 hours per week.
- Schedule: Includes 4 hours of overlap with Pacific Time.
- Language: Professional reading and writing ability in Polish.
- Experience level: Entry level.
- You need strong analytical judgment, creative prompt-writing ability, clear written communication, and close attention to detail.
- A bachelor's 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 a related role is preferred.
- You need reliable computer equipment and internet access.
- You must be willing to use a primary personal Google account and relevant personal data sources for evaluation.
How It Works
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI Training Work
AI training is the human work behind systems that understand and generate text. Contributors write examples, compare model responses, and explain which answers are accurate, useful, and natural so AI systems can improve.
