Dutch Personalization AI Quality Analyst

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. It connects contributors with opportunities to help shape modern AI, build a credible work profile, and grow experience in a fast-expanding field.

Creating an OpenTrain account is free, and you can apply in minutes. As an OpenTrain contractor, you will contribute to specialized AI quality work while developing experience for your long-term portfolio.

• Remote contract work available worldwide
• Part-time schedule of 20+ hours per week
• $20 per hour

About AI Personalization Evaluation

AI training is the human side of building artificial intelligence. People review model outputs, write and assess prompts, and provide detailed feedback that helps AI systems become more accurate, useful, and natural.

In this role, you will evaluate whether an AI assistant uses personal context responsibly and effectively. Your judgments will help distinguish genuinely helpful personalization from unsupported claims, awkward connections, hallucinations, or unnecessary explanation.

• Work with conversational AI and personalization features
• Assess relevance, helpfulness, naturalness, and evidence-based reasoning
• Contribute to the development of more capable AI systems

The Role

OpenTrain is recruiting a Dutch Personalization AI Quality Analyst for a Dutch-language evaluation project. You will create realistic prompts based on your own experiences and review how an AI assistant uses information from prior conversations and connected activity.

The work combines Dutch language expertise, creative multi-turn prompt design, and careful model evaluation. You will assess whether personalization is grounded in available evidence, integrated naturally, and genuinely useful rather than forced, inaccurate, or overexplained.

• Experience level: Entry level
• Employment type: Part-time contractor
• Time requirement: 20+ hours per week
• Language: Dutch reading and writing proficiency
• Pay: $20 per hour
• Work location: Worldwide and remote

What You'll Do

You will evaluate conversations that typically contain one to five turns. Each task requires close attention to the user's intent, the available personal context, and the assistant's specific wording and conclusions.

After reviewing each conversation, you will document clear, defensible reasoning. You will also delete evaluation conversations afterward to preserve clean personal chat history and maintain data hygiene.

• Design and execute creative multi-turn conversational prompts grounded in personal context
• Test how effectively an assistant uses personal information and experiences
• Evaluate responses against the intent of each starting prompt
• Identify unsupported claims, flawed inferences, hallucinations, and incorrect personalization
• Assess whether personal information is incorporated naturally without robotic narration
• Compare and rank two responses side by side
• Judge helpfulness, usability, enjoyment, and overall quality
• Write concise, structured rationales that reference specific conversation turns
• Delete evaluation conversations after completing reviews

Requirements

You should be comfortable reviewing nuanced or ambiguous AI responses and making consistent judgments when two answers have competing strengths. Strong Dutch reading and writing ability is required for the Dutch-language project.

The work also requires careful handling of personal context. You must be willing to use a primary personal Google account and follow the project's data hygiene expectations, including deleting evaluation conversations afterward.

• Strong Dutch reading and writing ability
• Analytical judgment when reviewing nuanced, ambiguous, or competing responses
• Ability to create creative multi-turn prompts from personal context
• Understanding of grounding, natural integration, incorrect inferences, forced connections, and hallucinations
• Excellent written communication and structured annotation skills
• Careful attention to subtle differences between side-by-side responses
• Ability to identify overnarrating and unnatural phrasing
• Willingness to use a primary personal Google account
• Reliable computer access and a strong internet connection

Helpful Background

A bachelor's degree or equivalent experience in policy, law, ethics, linguistics, journalism, computer science, or another analytical field is helpful. Previous experience in data annotation, AI quality evaluation, content moderation, or a related area is preferred.

This is an entry-level opportunity, so the most important qualities are strong Dutch communication, thoughtful analysis, creativity, and the ability to explain decisions clearly.

• Policy, law, ethics, linguistics, journalism, or computer science background
• Previous data annotation or AI quality evaluation experience
• Content moderation experience
• Independent remote work experience

Why This Work Matters

Modern AI assistants learn from examples and evaluations prepared by people. By testing realistic conversations and identifying subtle failures in personalization, you will help make AI interactions more relevant, accurate, and genuinely useful.

AI training and data-labeling work is a flexible way to participate in cutting-edge technology. Many contributors work remotely and choose projects that fit around their existing commitments while building specialized experience.

• Shape how conversational AI understands and uses personal context
• Work remotely on a fast-growing area of AI development
• Build experience in prompt design and model response evaluation
• Develop a portfolio of specialized AI training work

How to Apply

Create a free OpenTrain account and apply in minutes. Highlight your Dutch proficiency, analytical judgment, prompt-writing ability, and any relevant experience with AI evaluation, annotation, moderation, or independent remote work.

If selected, you will complete work as an OpenTrain contractor under the project's requirements, including the 20+ hour weekly commitment and personal-account data hygiene expectations.

• Create or update your free OpenTrain profile
• Show your Dutch reading and writing proficiency
• Describe relevant evaluation, annotation, or analytical experience
• Apply for the Dutch Personalization AI Quality Analyst role