Mathematics & Python Expert – Freelance AI Trainer

Mathematics & Python Expert – Freelance AI Trainer (Remote)

Mindrift connects specialists with project-based AI opportunities for leading tech companies, focused on testing, evaluating, and improving AI systems. Please note that this is project-based work, not permanent employment.

About the Role

As a Mathematics & Python Expert, you will design challenging mathematics problems to test frontier AI models. These problems must have answers verifiable by code and require specialized tools beyond standard libraries like NumPy or SymPy. You will work within a sealed Linux container environment where a programmatic judge evaluates the AI's responses.

Key Responsibilities

• Select an anchor tool (e.g., Z3, cvc5, SageMath) and design problems that specifically leverage its capabilities.
• Write Python reference solutions for the designed problems.
• Optionally supply necessary input files.
• Determine the numerical answer and the acceptable margin of error for the AI's solution.
• Test problems against the AI model, tuning difficulty to achieve a success rate between 10-30%.
• Iterate on problem design based on AI performance and feedback from senior reviewers to ensure high task quality.
• Gain a deeper understanding of specialized mathematical tools and frontier AI model behavior.

Requirements

• Degree in Mathematics (Pure or Applied) or a related field.
• Minimum of 2 years of experience in research, applied work, or teaching.
• Proficiency in Python for writing reference solutions.
• Fluency with, or a strong willingness to independently learn, at least one scriptable mathematical package such as Z3, cvc5, Macaulay2, Singular, CasADi, IPOPT, SDPB, G6K, fpylll, or SageMath.
• Ability to design problems that genuinely necessitate the use of a specialized solver.
• Strong written English skills (C1+).
• Experience with AI or machine learning is a plus.

Note: If you don't have prior experience with the listed tools but are eager to learn and apply them quickly, you are still encouraged to apply.

What We Offer

• Compensation: Earn up to $45 per hour equivalent, based on your expertise and contribution pace. Compensation may vary across different projects.
• Work Model: Fully remote freelance opportunities.
• Project Focus: Engage in challenging, project-based work with leading tech companies.
• Skill Development: Enhance your expertise in specialized mathematical tools and gain practical insights into AI model capabilities.
• Flexibility: Contribute to projects estimated at 10-20 hours per week during active phases.

How It Works

1. Apply
2. Pass qualification(s)
3. Join a project
4. Complete tasks
5. Get paid