Life Sciences Scientific Reasoning Expert

Create advanced, non-code life sciences reasoning tasks and expert reference answers for AI evaluation. This remote contract role requires strong biological research expertise and at least 20 hours per week.

100% Remote

Contract, part-time

Engagement

Remote

Location

Oct 20, 2026

Posted

You will create challenging scientific reasoning tasks based on realistic biological research scenarios. Your work will use evidence such as experimental findings, datasets, figures, sequences, structures, and literature-style material to test whether AI systems can interpret and reason about science.

The role focuses on deep biological judgment rather than programming or simple factual recall. You will also review work for accuracy, clarity, and reproducibility.

• Design original life sciences problems that require advanced interpretation and scientific reasoning.
• Write precise questions and rigorous reference answers with objectively defensible conclusions.
• Check biological claims, calculations, assumptions, and conclusions for scientific accuracy.
• Find ambiguity, alternative valid interpretations, and questions that lack enough constraints.
• Document tasks clearly, incorporate reviewer feedback, and maintain quality.
• Create problems that distinguish AI models by their ability to reason from scientific evidence.

• Pay:
  The source does not provide a rate.
• Time: 20 or more hours per week.
• Location:
  
  Remote and open worldwide.
• Language:
  Professional fluency in English.
• Education or experience: A Ph.D., postdoctoral experience, or equivalent advanced research experience in the life sciences.
• Subject expertise:
  Strong knowledge of at least one area such as molecular biology, genetics and genomics, cell biology, biochemistry, neuroscience, microbiology, immunology, ecology, structural biology, biophysics, developmental biology, cancer biology, or pharmacology.
• Core skill:
  Ability to interpret scientific experiments, biological data, and research findings.
• Communication skill:
  Ability to turn complex scientific concepts into precise, self-contained problems with defensible answers.
• Helpful background:
  Graduate-level assessment design, manuscript or research review, evaluation of AI-generated scientific responses, or work with AI and machine learning evaluation datasets.
• Also valued:
  Publications in peer-reviewed journals or conferences, and graduate-level teaching or mentoring experience.
• Create demanding questions, review AI-generated responses, and explain errors using authoritative sources. This part-time contract pays $30 to $60 per hour equivalent, paid per task that meets the project specifications.
• Review AI conversations, identify missing context, and assess whether recommendations use enough information. This remote project runs for about three weeks, requires 20+ hours per week, and pays per task.
• Remote
  · Albania, Algeria, Angola +128 more
• Create research-backed questions and answers, test AI model responses, and improve evaluation content remotely for $20 to $60 per hour equivalent, paid per task.
• Create difficult, evidence-based question-and-answer materials for advanced AI model evaluation. Earn $30 to $60 per hour equivalent, paid per task that meets the project specifications, with fewer than 20 hours per week.

• Experience:
  
  Entry level

AI training is the human work behind systems that learn from examples, including writing and reviewing model responses and evaluating scientific reasoning. People with deep subject knowledge help create reliable examples and judgments that improve how these systems handle specialized questions.