Statistician & Experimentation Expert AI Evaluation

You will create and evaluate AI training data focused on causal inference and A/B testing. The work tests whether AI models can reason correctly about study design, statistical analysis, and results.

You will use expert statistical judgment to assess model outputs, explain sound reasoning, and identify conclusions that appear rigorous but are not valid.

• Write realistic statistics scenarios about study design, analysis, and interpretation.
• Create reference answers that explain the reasoning behind modeling choices.
• Evaluate AI outputs for methodological soundness and accurate interpretation.
• Identify analyses that use an apparently rigorous approach but draw invalid conclusions.
• Apply expertise in experimental design, causal inference, and A/B testing.

This is an ongoing, fully remote independent contractor role. The schedule is asynchronous, with flexibility to adjust your workload from week to week.

The role requires advanced statistical training and applied experience, along with the ability to explain technical decisions clearly.

• Work at least 10 hours per week, with no maximum.
• Open to candidates in the countries listed for this role.
• Fluent in English.
• Hold an advanced degree in statistics, biostatistics, or a closely related quantitative field.
• Have at least two years of applied statistics or experimentation experience.
• Have strong knowledge of experimental design, causal inference, or A/B testing.
• Be fluent in R or Python for statistical analysis.
• Document and explain modeling decisions clearly.
• Experience with large-scale experimentation platforms, Bayesian methods, instrumental variables, or synthetic control methods is helpful.
• A PhD in statistics, biostatistics, or econometrics is relevant background.
• Payment is made weekly via Stripe.

Apply on OpenTrain with your resume, then complete the application on the hiring site. OpenTrain helps you find this AI training role and begin the application, but does not employ or pay contractors for it.

AI training work uses examples and expert reviews to help artificial intelligence systems produce better results. Statistics experts are important when models must reason about evidence, experiments, uncertainty, and cause-and-effect relationships.