This opportunity involves reviewing a mix of customer service-style responses and technical content, with a strong focus on coding knowledge, JSON structure, formatting accuracy, and overall response quality. Experts should be comfortable identifying whether outputs are clear, helpful, natural, technically accurate, and aligned with project guidelines.

Ideal experts will have strong attention to detail, strong written communication skills, and the ability to evaluate both the customer experience and the technical correctness of a response. A solid understanding of JSON is especially important, including the ability to read, validate, troubleshoot, and identify structural or formatting issues.

We are especially sourcing experts with fluency in one or more of the following language varieties:

Dutch, as spoken in the Netherlands.

Welsh, as spoken in the United Kingdom.

French, as spoken in Canada.

German, as spoken in Germany.

Italian, as spoken in Italy.

Portuguese, as spoken in Brazil.

This opportunity is best suited for experts who can combine language fluency, customer service judgment, and technical evaluation skills to help improve the quality, accuracy, and usability of AI-generated outputs.

Please speak naturally and spontaneously, as if responding in real time. Do not pre-plan, write, or script your response—what we are evaluating is natural voice cadence. Responses may be up to one minute long; this may be a few sentences, depending on what feels natural.

GDM Audio Evals.
This verification takes approx. 20 mins.

Work arrangement: Fully remote

Applicant eligibility: 145 eligible countries

• Fees vary based on service and complexity.