Data Science Expert - AI Model Evaluation

Description

You can get paid for applying your data science skills to help improve AI.

The company is a Silicon Valley firm working on the intelligence behind frontier AI. It is hiring experienced data science professionals for paid, project-based work that trains and evaluates the next generation of AI systems.

You work remotely and asynchronously, on projects that fit around your current schedule.

Your expertise helps AI systems learn what accurate, rigorous, and professionally sound data science looks like in practice.

Responsibilities

• Depending on the project, your tasks may include the following.
• Analyze datasets, statistical analyses, experiments, models, metrics, and data-driven conclusions.
• Review AI-generated data science solutions, including analyses, code, models, explanations, and reasoning.
• Judge whether AI-generated outputs are statistically sound, technically accurate, logically consistent, and backed by the available data.
• Examine data pipelines, feature engineering, model selection, statistical methods, experiments, visualizations, and machine learning approaches.
• Work through realistic data science problems and explain how an experienced data scientist would tackle them.
• Compare several AI-generated solutions and identify meaningful differences in accuracy, method, reasoning, and professional judgment.
• Write challenging data science tasks and the criteria used to measure AI systems.
• Point out errors, flawed assumptions, data quality problems, methodological weaknesses, and unsupported conclusions in AI-generated work.
• Projects may cover statistical analysis, machine learning, predictive modeling, experimentation, data visualization, feature engineering, forecasting, causal inference, data analytics, and other areas that match your expertise.

Requirements

• The company is looking for experienced data science professionals, though you do not need to be an AI researcher.
• You need at least 3 years of professional experience in data science, statistics, analytics, machine learning, or a closely related field.
• You need strong statistical, analytical, and quantitative reasoning skills.
• You need hands-on experience with real-world datasets and data-driven problems.
• You must be able to critically evaluate data science analyses, models, and conclusions.
• You need a sharp eye for detail and the ability to spot errors, inconsistencies, flawed assumptions, and methodological issues.
• You need sound professional judgment.
• You must be able to explain clearly why a data science approach or conclusion is right or wrong.
• Experience in the following areas is valuable: statistical analysis, machine learning, predictive modeling, data visualization, experimental design and A/B testing, feature engineering, time series and forecasting, causal inference, and data analytics.
• Experience with Python, R, SQL, or similar tools is valuable.
• Experience in model evaluation and validation is valuable.
• Experience in data quality and preprocessing is valuable.
• Prior experience working in AI is not required.
• Knowing how to do data science well matters most.
• Typical tasks include reviewing an AI-generated analysis and listing its problems, checking whether a statistical analysis is correct and supported by the data, judging whether a machine learning approach suits a problem, comparing two AI-generated solutions to pick the stronger one, and reviewing an AI-generated model evaluation for methodological flaws.
• Other typical tasks include working through a realistic dataset or scenario and explaining your reasoning, finding data quality issues, statistical errors, flawed assumptions, or unsupported conclusions, checking whether an AI-generated visualization or interpretation fairly reflects the data, and building hard data science problems used to test AI systems.
• The aim is not just to check whether data science work looks reasonable. The aim is to capture the judgment an experienced data scientist uses to decide whether an analysis truly holds up.

Conditions

The work is paid on a project basis.

Rates depend on your experience, specialization, and the project requirements. Your rate and project scope are shown before you commit to anything.

Some projects are short. Others may recur over a longer period.

There is no fixed schedule and no minimum commitment.

The work is remote, asynchronous, and project-based, with flexible hours that can fit around your existing job or other commitments.

Pay applies to qualifying project work only.

A plausible-looking answer is not enough. The AI needs to learn how experienced data scientists analyze and interpret data, choose statistical methods, evaluate model performance, identify bias and data quality issues, design and evaluate experiments, assess assumptions and limitations, interpret results correctly, compare modeling approaches, recognize important edge cases, catch errors and inconsistencies, explain conclusions clearly, and decide when an analysis is reliable.

The process starts with a short assessment matched to the type of data science work you would do. When your experience fits an active project, you receive the scope, requirements, and compensation. You then complete the project remotely and asynchronously and are paid the agreed rate for finished work.

Benefits

• You are paid for your expertise while working on your own schedule.
• You help shape how AI learns to reason about data.