Data Science AI Trainer

When AI gets the details wrong in your field, real people pay for it. The people who know the work best have to be the ones training these models. You're not being replaced by AI. You're making sure it's used accurately and safely.

• Design challenging, realistic data science tasks based on professional workflows, including scenarios, colleague‑style prompts, and supporting materials such as datasets, schemas, notebooks, metric definitions, and stakeholder briefs.
• Create or anonymize realistic professional files needed to complete each task.
• Run tasks through frontier AI models and evaluate the resulting analyses, code, models, SQL queries, and written deliverables against professional standards.
• Compare model outputs produced from identical prompts and files, determine which is stronger, and document where each response falls short.
• Write detailed grading rubrics and tests defining what a correct deliverable must include, such as appropriate data‑quality handling, statistical tests, validation methods, and interpretations.
• Identify and document concrete failures, including data leakage, confounding, incorrect joins, misapplied statistical tests, fabricated or ignored data, code that does not meet the brief, and unsupported interpretations.
• Contribute expertise across analytics, machine learning, experimentation, forecasting, and data‑engineering‑adjacent work.
• Review and refine tasks developed by other subject‑matter experts.
• Propose, build, and run complete task scenarios based on real professional experience, incorporating expert‑review feedback and revisions.

• 2+ years of professional applied data science experience preferred in analytics, machine learning, statistical modeling, experimentation, analysis‑adjacent data engineering, or research; candidates with less experience may be considered if their practical work is substantial.
• Bachelor's degree in progress or completed in a quantitative field; advanced degrees are a plus but not required.
• Working proficiency in Python and/or SQL, including the ability to write, debug, read, and explain analysis code.
• Depth in at least one area: product or business analytics and BI, applied machine learning, statistics and experimentation, forecasting, analysis‑adjacent data engineering, or applied NLP/computer vision with shipped work.
• Experience independently owning a multi‑step analysis from raw or messy data through cleaning, modeling, interpretation, and decision‑making.
• Ability to evaluate whether results are valid and explain issues such as leakage, confounding, selection effects, multiple testing, power, and business or scientific interpretation.
• Current or recent hands‑on individual‑contributor experience; candidates whose experience is solely managerial are not a fit.
• Ability to draw on real‑world workflows and professional experience to create scenarios that test whether AI can perform the work correctly.
• Full professional or native‑level written and spoken English, with the ability to clearly explain complex reasoning and why an output is incorrect.
• Strong attention to detail, comfort with ambiguity, and the ability to verify claims against underlying data, sources, or facts.
• Ability to independently interpret feedback, determine what is correct, and apply it without extensive direction.
• Ability to ramp quickly on unfamiliar work using written instructions and incomplete source material.
• General familiarity with AI and LLM tools such as Claude or ChatGPT and the judgment to distinguish sound reasoning from plausible but incorrect answers.
• Comfort using cloud file tools, browser profiles, desktop applications, and everyday file‑management workflows.
• Availability for at least 10 hours per week, with consistent availability…