Data Scientist, AI Model Evaluation; Remote

20 or more hours a week, you choose when

Work on the cutting edge of AI, on your own schedule, and get paid for the data science judgment you already have.

The AI models changing how technical work gets done are trained and tested by real data scientists. The teams building them need experts who can create demanding prompts and datasets, judge whether a model's answer is correct and complete, write the reference answer, and explain where its methods or reasoning break down. Your judgment helps shape the next generation of models and makes their results more reliable.

This is that work, done remotely around your current job.

• Review model answers for correctness, completeness and sound reasoning across data science and statistics
• Create expert prompts, datasets and reference solutions for machine learning and experimentation
• Identify flawed methods, statistical errors and weaknesses in quantitative analyses
• Explain clearly why an approach is correct, incomplete or unsound

• Get paid $245–$280 an hour to build real AI skills, on top of your current job
• Work directly with frontier AI tools and see how they are built, tested and improved, experience most data scientists will not have for years
• Build a career edge for what comes next: AI is reshaping data science, and the people who understand it from the inside will lead that change
• Fully remote and flexible, no fixed shifts, no commute

• At least one year of professional experience at a leading company in technology, finance or research, including experience from the past seven years
• An undergraduate degree from a top-ranked university is preferred
• Strong knowledge of data science, statistics, machine learning, experimentation and quantitative reasoning, with clear English communication and the ability to work independently
• No experience in AI training work needed. You bring the expertise; we bring the training on the tools.

• $245–$280 per hour, paid hourly as an independent contractor
• 20 or more hours a week, you choose when; fully remote, United States