Turing — AI Quality Analyst (English)

Official Turing job listing

Work: Remote, contractor

Rate: $30/hour; the top rate is $30/hour.

Hours: At least 4 hours/day, up to 40 hours/week, with 4 hours of PST overlap. TTuring

What you will do The project focuses on evaluating Gemini's personalization capabilities. You will:

- Create 1–5 turn conversational prompts based on personal experiences/context.
- Test whether Gemini correctly uses information from sources such as past Gemini conversations, Gmail, Google Search and YouTube.
- Evaluate responses for:
- Grounding — are claims about the user actually supported?
- Integration — is personal information incorporated naturally?
- Helpfulness — is the response practical and relevant?
- Naturalness and excessive/forced personalization.
- Compare two AI responses side-by-side (SxS) and rank which is better.
- Write clear rationales explaining why one response is better, referring to specific conversation turns.
- Check model/debug information to verify that the appropriate personal data sources were used.




- Maintain data hygiene by deleting evaluation conversations afterward. TTuring+1

Requirements Turing is looking for someone with:
- Excellent English reading and writing ability.
- Strong analytical and critical-thinking skills.
- Ability to create creative, context-rich prompts.
- Ability to recognize hallucinations, incorrect personalization, unsupported inferences and forced connections.
- Excellent attention to detail.
- Strong written communication and ability to provide structured feedback.
- Ability to work independently remotely.
- A reliable laptop/desktop and internet connection.
- BS/BA or equivalent experience in areas such as Computer Science, Linguistics, Journalism, Ethics, Law, Policy, or another analytical field.
- Previous experience in AI evaluation, data annotation, content moderation, or similar work is preferred, not necessarily mandatory. TTuring

📌 AI Quality Analyst (English) (India)
🏢 Innova ESI
📍 India