AI Quality Analyst Gemini Personalization

As an AI Quality Analyst, you will evaluate a Gemini personalization feature. You will test how well the model uses information from past conversations, Gmail, Google Search, and YouTube activity to make its answers more relevant and helpful.

• Create and run multi-turn conversations, usually lasting one to five turns, using personal information and experiences.
• Check whether each response understands your intent and uses personalization in an appropriate way.
• Review responses for grounding issues, including claims that are unsupported, based on flawed inferences, or made up.
• Assess whether personal information is included naturally instead of being repeated in a robotic way.
• Compare and rank two model responses based on overall helpfulness, ease of use, and quality of the experience.
• Write clear reasons for your rankings and refer to specific conversation turns.

What it pays and takes

This is a three-month contractor engagement. The listing does not provide a pay rate.

• Time: At least 20 hours per week, with at least 4 hours per day and up to 40 hours per week.
• Schedule: Includes 4 hours of overlap with Pacific Standard Time.
• Location: Worldwide.
• Language: Strong English reading and writing skills are required.
• Equipment: A desktop or laptop and a reliable internet connection.
• Skills: Strong analytical thinking, careful attention to detail, creative prompt design, knowledge of personalization concepts, and excellent written communication.
• Helpful background: A bachelor's degree or equivalent in policy, law, ethics, linguistics, journalism, computer science, or another analytical field.
• Helpful experience: Data annotation, AI quality evaluation, content moderation, or related work.

How it works

Apply on OpenTrain with your resume, then complete the application on the hiring site.

About AI training work

AI training work uses human-written examples, ratings, and reviews to help improve how AI systems respond. People with strong judgment and attention to detail are needed to assess whether model outputs are accurate, useful, and appropriate.