Vietnamese AI Quality Analyst Personalization

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. OpenTrain AI is hiring and contracting for this remote opportunity, helping contributors discover meaningful AI work and grow their experience in a fast-moving field.

• Free to create an OpenTrain account
• Remote contractor opportunity
• Work focused on Vietnamese-language AI evaluation

About AI Training Work

AI training is the human side of building artificial intelligence. People review and evaluate model outputs so AI systems can become more accurate, relevant, and helpful. This work gives contributors a direct role in shaping how modern AI behaves while offering flexible opportunities that can fit around other commitments.

• Contribute to the development of cutting-edge AI systems
• Use language skills, analytical judgment, and attention to detail
• Work remotely with a computer and reliable internet connection

The Role

OpenTrain is recruiting an AI Quality Analyst to evaluate a new personalization feature for Gemini. You will assess how effectively the model uses personal data from past conversations, Gmail, Search, and YouTube to produce more relevant and helpful responses.

This contractor role focuses on Vietnamese-language evaluation. You will create prompts from your own experiences, review personalized outputs, and assess them for Grounding, Integration, and Helpfulness.

• Language focus: Vietnamese reading and writing
• Experience level: Entry level
• Pay: $15 per hour
• Engagement: 3 months
• Work arrangement: Fully remote

What You'll Do

You will test personalized AI behavior through realistic, creative conversations and provide structured judgments about the quality of each response. Your evaluations should reference specific conversation turns and clearly explain why one output is stronger than another.

Maintaining data hygiene is an important part of the role. Evaluation conversations must be deleted after use.

• Design and execute multi-turn conversational prompts spanning 1 to 5 turns
• Use personal information and experiences as context for prompts
• Check responses for Grounding issues, including unsupported claims and hallucinations
• Assess whether personal data is integrated naturally into the response
• Rank two model responses side by side based on helpfulness and enjoyment
• Write clear, structured rationales for your rankings
• Delete evaluation conversations after use

Requirements

This role requires strong Vietnamese reading and writing ability, careful analysis of nuanced personalization, and the creativity to design prompts based on personal context. You should be comfortable making independent judgments about subtle differences in AI response quality and explaining those judgments in writing.

A desktop or laptop with reliable internet is required. The role is remote and requires flexibility to support global 24-hour operations.

• High-level Vietnamese reading and writing proficiency
• Experience evaluating AI model responses for personalization quality
• Ability to create creative multi-turn prompts using personal context
• Skill in side-by-side model response ranking and rationale writing
• Strong analytical thinking about Grounding, Integration, and Helpfulness
• Exceptional attention to detail
• Excellent written communication
• Self-motivation and ability to work independently
• Desktop or laptop with reliable internet

Helpful Background

Experience in data annotation, AI quality evaluation, or content moderation can be helpful. A BS, BA, or equivalent experience in Policy, Law, Ethics, Linguistics, Journalism, Computer Science, or a related field is also helpful, but the listed experience level for this opportunity is entry level.

• Data annotation experience
• AI quality evaluation experience
• Content moderation experience
• Background in Policy, Law, Ethics, Linguistics, Journalism, Computer Science, or a related field

Schedule and Commitment

The engagement requires at least 4 hours per day and supports up to 40 hours per week. The listed time requirement is 20 or more hours per week, with 4 hours of overlap with Pacific Standard Time required.

• Minimum commitment: 4 hours per day
• Listed time requirement: 20+ hours per week
• Maximum schedule: up to 40 hours per week
• Required overlap: 4 hours with PST
• Duration: 3-month engagement
• Location: Worldwide and fully remote