Junior Data Scientist, Ai Evaluation

The Work
You will assess AI-generated answers to problems in data science, statistics, machine learning, experimentation, and quantitative reasoning. Your reviews will help identify weak methods, statistical errors, and unsupported conclusions.You will also create prompts and datasets that test analytical reasoning, develop reference solutions, and write structured feedback to improve model accuracy and usefulness.Evaluate answers to data science, statistics, machine learning, experimentation, and quantitative problems.Create prompts that test analytical and quantitative reasoning.Develop datasets and reference solutions for model evaluation.Find flawed methods and statistical errors in AI-generated work.Explain technical and statistical issues in clear written feedback.What It Pays and Takes
This is a remote, independent contractor role listed as entry level. It is available on a part-time basis and focuses on expert review, prompt and dataset creation, reference-solution development, and structured feedback.Pay: $150 to $200 per hour.Work arrangement: Remote, part-time independent contractor.Location: Open to candidates in Canada, the United Kingdom, and the United States.Language: Fluent written English.Experience: At least one year of professional experience at a leading technology, finance, or research company, including some experience within the past three years.Knowledge: Strong understanding of statistics, machine learning, experimentation, and quantitative reasoning.You must be able to judge whether analytical methods and conclusions are technically sound.Helpful experience includes data analysis, data modeling, Python, Pandas, or related quantitative workflows.About AI Training Work
AI training is the human work behind many AI systems. People review model answers, create examples, and give feedback so models produce more accurate and useful results. Experience in fields such as statistics, machine learning, and data science is valuable because it helps reviewers judge technical reasoning carefully.
#J-18808-Ljbffr