Biology AI Benchmark Designer for Real-World Data

OpenTrain AI, Inc.; Northern (KY); Hybrid; USD 90,000 - 130,000; Full time

Job summary

OpenTrain is seeking a Life Sciences AI Evaluation Task Designer to create challenging benchmark tasks that help evaluate how AI systems reason about real scientific work. You will use authentic research artifacts and technical challenges to represent advanced life sciences practice, including experimental design critique, protocol analysis, primary data interpretation, and methodological judgment.

You will source and synthesize data, define the correct methodological approach, and build

Qualifications

• Bachelor's or master's in biology, biomedical sciences, biochemistry, or closely related field.
• At least 2 years of hands-on research with primary data analysis.
• Experience with protocol-driven, multi-source scientific workflows and documenting results.
• Strong written English and ability to evaluate scientific literature.

Responsibilities

• Develop complex research evaluation tasks in biology or related life sciences.
• Source and synthesize raw experimental data, protocols, literature, and supplementary documents.
• Define the correct methodological approach and document the biological interpretation for each task.
• Build multi-item rubrics assessing rigor, data interpretation, and fidelity to sources.
• Ensure task scenarios capture ambiguity, nuance, and depth of real-world science.
• Collaborate asynchronously with coordinators to refine deliverables and maintain objective evaluation standards.

Skills

Scientific reasoning

Source synthesis

Methodological evaluation

Written English

Data interpretation

Education

Bachelor's or Master’s in Life Sciences

Job description

OpenTrain is seeking a Life Sciences AI Evaluation Task Designer to create challenging benchmark tasks that help evaluate how AI systems reason about real scientific work. You will use authentic research artifacts and technical challenges to represent advanced life sciences practice, including experimental design critique, protocol analysis, primary data interpretation, and methodological judgment.

You will source and synthesize data, define the correct methodological approach, and build