STEM PhD Expert for AI Reasoning & Evaluation
$50–$150 / hourRemote · AR, AU, CA +
- Machine Learning
- Engineering
- Physics
We're hiring PhD-level experts to help train and evaluate advanced AI models. This is remote, flexible contract work where your academic expertise directly shapes how cutting-edge models reason, answer, and improve. You'll review AI-generated content in your field, identify where the model gets things right or wrong, and help raise the quality bar on technical reasoning.
This is a short-term engagement running through the end of June, with potential to extend depending on project needs.
Key Responsibilities
You may contribute your expertise by:
- Assessing the factuality and relevance of domain-specific text produced by AI models
- Crafting and answering questions related to Machine Learning and AI
- Evaluating and ranking domain-specific responses generated by AI models
What We're Looking For
- A PhD (completed or in final stages) in one of the following fields: Machine Learning / AI, Computer Science, Engineering, Statistics, or a closely related quantitative subdomain such as Mathematics or Physics
- Strong analytical and critical-thinking skills, with the ability to spot subtle errors in technical reasoning
- Fluent written English and the ability to communicate complex ideas clearly
Nice to Have
- Research experience (academic or industry)
- Prior experience with data annotation or AI model evaluation
- Experience reviewing or publishing research papers
Compensation
Up to $150/hr, depending on your area of expertise, depth of experience, and assessment performance. Location
This role is fully remote. We are currently accepting applicants based in:
United States, Canada, Puerto Rico, Mexico, United Kingdom, Australia, New Zealand, and Argentina.
Description from Braintrust. Full details and the application are on Braintrust's site.
About Braintrust
Braintrust is a talent network where companies hire freelancers directly. Keywork lists the AI training work posted there by AI data companies such as DATAmundi, Outlier and Toloka: language data, expert evaluation and voice work.
All Braintrust roles →