Dutch Language AI Quality Trainer & Assessor (209370)

Overview

Part time: Dutch Language AI Quality Trainer & Assessor

Use your fluent Dutch & English skills and join a forward-thinking team at the forefront of artificial intelligence, who are partnering with Aquent to shape the future of personalized digital experiences.

In this pivotal role, on a 4 week project, you will directly influence the quality and relevance of a leading AI model, ensuring it truly understands and serves users on a deeply personal level. Your insights will be critical in advancing the next generation of AI capabilities, making a tangible impact on how millions interact with technology every day.

We are seeking a fluent Dutch & English speaking, highly analytical and creative individual to join this project.

This is an extraordinary opportunity to be at the heart of innovation, contributing to the development of a cutting-edge personalization feature for a widely used AI model.

You will leverage your own experience, unique perspective and sharp analytical skills to meticulously evaluate how the model integrates personal context, directly enhancing its helpfulness and relevance for users worldwide.

**Responsibilities:**
* Evaluate a new personalization feature for a leading AI model in Dutch language.
* Assess the model’s ability to utilize information from past AI model conversations, email, search history, and video activity to generate relevant and helpful responses.
* Actively design creative, multi-turn prompts based on your personal experiences to thoroughly test the model’s capabilities.
* Use analytical skills to assess the quality of the model’s personalized responses, evaluating dimensions like Grounding, Integration, and Helpfulness.
* Contribute to the success of key initiatives by identifying incorrect personalization, poor inferences, and forced connections within the model’s outputs.

**Must-Have Qualifications:**
* At least 6 months of experience regularly using a Gemini and willingness to use your experience and activity for assessment purposes.
* A solid history of searches in the Dutch language.
* You must be based in The Netherlands or Belgium .
* Native speaker or fully fluent in Dutch, with business-level proficiency in English (instructions and project communications are in English).
* Part-time commitment: You can commit to minimum 20 hours per week, with at least 4 hours per day (Monday – Friday), for a period of 4 weeks.
* Work hours must have at least 2 hours of daily overlap with US Pacific (West Coast) time for meetings; eg early evening work will be required.
* Degree Qualified.
* Exceptional Analytical Thinking: Demonstrated ability to evaluate nuanced and ambiguous AI responses, specifically assessing personalization quality.
* Creative Prompt Engineering: Proven experience in designing creative, multi-turn starting prompts based on personal context to thoroughly test model capabilities.
* Strong Evaluation Acumen: Deep understanding of personalization concepts, including the ability to identify incorrect personalization, poor inferences, and forced connections.

**Nice-to-Have Qualifications:**
* Experience in data annotation, AI quality evaluation, content moderation, or a related role is strongly preferred (but not essential if proven on assessment).