Behavioral Scientist - AI Trainer

About the role

This position is for quantitative specialists who want to contribute to AI training. You will collaborate with advanced AI systems by reviewing model outputs, tackling technical challenges, and giving structured feedback that improves how these models handle data, modelling, and scientific reasoning.

The work suits people from a wide range of quantitative backgrounds, including data science, astrophysics, economics, biostatistics, operations research, and similar fields. If your approach is grounded in rigorous analysis of data and models, your expertise is relevant here. Some contributors combine this work with a full-time job, while others make it their main focus.

To begin, you will create an account and complete a short assessment that functions as the screening step. Candidates who pass receive a confirmation email, after which paid assignments become available on the platform.

Advantages of contracting

• You can select the projects you want to take on and decide when to work on them.
• The role offers schedule flexibility and can be done on your own computer from home.
• Compensation is hourly, starting at USD 50-100+ per hour, with some projects offering bonus pay.

Responsibilities

• Review AI-produced quantitative outputs for statistical soundness, modeling quality, scientific logic, and practical accuracy.
• Create and solve quantitative tasks used to train and evaluate AI systems, including forecasting, experimental analysis, optimization, and statistical inference.
• Produce concise technical write-ups and maintain clearly documented analytical code.
• Share feedback that helps improve the next generation of AI models focused on quantitative reasoning.

Qualifications

• At least 2 years of practical experience in a quantitative role or research setting such as data science, statistics, economics, finance, physics, biology, epidemiology, operations research, or a related area.
• Basic coding ability is required, along with confidence in writing and reviewing analytical code from start to finish.
• Hands-on familiarity with statistical techniques, predictive modeling, and experiment design such as A/B testing, hypothesis testing, regression, classification, and time-series forecasting.
• Strong English fluency at native or bilingual level, plus excellent writing skills.
• A bachelor's degree in a quantitative discipline is preferred, such as Statistics, Computer Science, Mathematics, Engineering, or a similar subject; a master's degree or PhD is an advantage.
• Additional proof of expertise is beneficial, such as a Kaggle ranking, AWS/GCP ML certifications, or comparable credentials.

Additional information

Compensation is processed through PayPal. You will never be asked to pay any money as part of this role. PayPal will manage any currency conversion from USD. This is an independent contractor engagement, and only candidates based in the United States, Canada, the United Kingdom, Ireland, Australia, or New Zealand will be considered.