Quality Evaluator – Fluent English (Brazil)

Quality Evaluator – English

Help shape the future of AI by evaluating and comparing responses from advanced AI models.

• Location: Remote

• Type: Independent contractor / Part-time

• Commitment: up to 3 months

• Compensation: $7-$12/hour

About the role

At Anyone AI, we create high-quality training and evaluation data for frontier AI models.

We’re looking for English-fluent experts with strong writing skills, critical thinking, and extensive hands-on experience using large language models (LLMs).

In this role, you’ll interact directly with AI models and compare their responses side by side to determine which best meets the user’s needs. Your evaluations and written explanations will help improve the accuracy, helpfulness, and reliability of AI systems.

What you’ll do

• Interact with different AI models and tools across a wide range of prompts and topics.

• Compare two or more AI-generated responses side by side.

• Assess accuracy, helpfulness, instruction-following, clarity, and overall response quality.

• Follow detailed rating guidelines and multi-step evaluation workflows.

• Write clear, well-reasoned explanations supporting your evaluation decisions.

• Apply evaluation criteria consistently across tasks.

What we’re looking for

• Fluency in English, with strong reading comprehension and writing skills.

• Extensive hands-on experience using LLMs and critically assessing their outputs.

• Strong analytical judgment and the ability to weigh multiple dimensions of response quality.

• Excellent attention to detail and the ability to follow complex instructions carefully.

• Ability to explain evaluation decisions clearly and convincingly.

• Comfort switching between AI models, tools, and multi-step workflows.

Nice to have

• Previous experience in AI response evaluation, data annotation, or quality assurance.

• Experience in editing, research, fact-checking, or analytical writing.

• Familiarity with evaluation rubrics and side-by-side comparison tasks.