Portuguese Scientific AI Safety Evaluator

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We help people discover specialized projects, build a credible professional profile, and apply for opportunities that match their expertise.

Creating an OpenTrain account is free, and contributors can use their profile to grow a lasting portfolio in the rapidly expanding field of AI training.

About AI Safety Evaluation Work

AI training is the human side of building artificial intelligence. Experts prepare examples, write prompts, review model responses, and provide structured feedback that helps AI systems become more accurate, useful, and safe.

This role focuses on advanced scientific topics in Portuguese, giving qualified researchers an opportunity to apply their subject-matter knowledge directly to cutting-edge AI development. Prior AI or machine learning experience is not required.

The Role

OpenTrain is seeking PhD-level scientists who are proficient in Portuguese and English to evaluate how AI models handle specialized technical and scientific topics. You will assess responses for scientific accuracy, helpfulness, and responsible treatment of sensitive or dual-use information.

This is remote contract work with an expected commitment of approximately 7 hours per week. Applicants may be based anywhere, with Portugal and Western Europe preferred.

• Pay: $50-$54 per hour
• Schedule: Approximately 7 hours per week
• Work arrangement: Remote, part-time contract
• Location: Worldwide applicants may be considered
• Languages: Native or near-native Portuguese and business-level written English

What You’ll Do

You will create expert-level prompts in Portuguese and apply structured evaluation guidelines to model outputs. Your work will help identify where scientific responses are accurate and useful, as well as where they are incomplete, misleading, or unsafe.

You will review technical content across specialized scientific subjects and classify prompts and conversations consistently. Sound judgment is especially important when handling dual-use information and other sensitive scientific material.

• Write expert-level scientific prompts in Portuguese
• Evaluate model responses for scientific accuracy and helpfulness
• Annotate responses according to structured guidelines
• Classify prompts and conversations
• Identify inaccurate, unsafe, or insufficiently helpful responses
• Assess the responsible handling of sensitive and dual-use information

Required Qualifications

You should have an ongoing or completed PhD in chemistry, biology, or a closely related scientific field. The role requires deep knowledge of a specialized subfield and familiarity with modern laboratory or computational techniques.

You must be able to evaluate technical content carefully in Portuguese and English, communicate clearly in written English, and make responsible judgments about scientific safety.

• Ongoing or completed PhD-level expertise in chemistry, biology, or a closely related field
• Deep knowledge of a specific scientific subfield
• Familiarity with modern laboratory or computational techniques
• Native or near-native Portuguese
• Business-level written English
• Ability to evaluate scientific accuracy and helpfulness
• Sound judgment regarding scientific safety and dual-use information

Helpful Scientific Background

Experience in a specialized scientific area can help you assess nuanced model responses and identify subtle accuracy or safety issues. Background in research, coursework, or technical content review is also relevant.

• Radiochemistry or nuclear chemistry
• Energetic materials or radiation physics
• Organic or inorganic synthesis
• Molecular biology, microbiology, virology, or genetics
• Chemical safety, biosafety, biosecurity, or hazardous-materials handling
• Reviewing, grading, or red-teaming technical content

Why Build an AI Training Career with OpenTrain

AI training and data labeling are expanding fields where specialists help shape how modern AI systems understand and respond to complex information. Remote, flexible projects can fit around research, teaching, or other professional commitments.

OpenTrain gives contributors a place to discover opportunities, demonstrate relevant experience, and build a professional portfolio in AI training over time.

• Apply scientific expertise to state-of-the-art AI systems
• Work remotely with a flexible part-time commitment
• Build experience in AI evaluation and safety
• Create a stronger professional portfolio through OpenTrain