This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for a Mathematics AI Training Expert - PhD based in India.

This is a highly specialized research-focused contract role where you will help advance the reasoning capabilities of frontier AI systems by designing rigorous and original mathematical problems. You will contribute directly to high-quality STEM training datasets used in AI model evaluation and improvement at cutting-edge research levels. The role involves creating deterministic, single-solution problems that challenge advanced mathematical reasoning rather than memorization. You will also develop complete, verified solutions with clear and precise explanations, ensuring reproducibility and technical accuracy. The environment is deeply analytical, detail-oriented, and research-driven, with strong emphasis on mathematical rigor and computational thinking. This is an opportunity to directly influence how next-generation AI systems learn to reason in complex mathematical domains.

• Design advanced and original mathematical problems for AI training and evaluation, ensuring they test deep reasoning rather than surface-level recall.
• Develop deterministic problems with a single verifiable correct answer, maintaining strict mathematical rigor and clarity.
• Produce complete, well-documented solutions with step-by-step reasoning and validated correctness.
• Apply computational tools where relevant, including Python and numerical libraries, to support problem construction and verification.
• Ensure all outputs are precise, reproducible, and clearly written in English for use in AI training pipelines.
• Contribute to improving the quality of AI reasoning datasets by identifying gaps and designing more challenging problem structures.
• Collaborate in a structured workflow focused on high-quality data generation for advanced model evaluation.

Requirements:

• Master’s or PhD in Mathematics or a closely related field such as applied mathematics, computational mathematics, or mathematical physics.
• Strong background in mathematical reasoning, proof-based problem solving, and/or research-level mathematical work.
• Proficiency in Python with experience using libraries such as NumPy, SciPy, Pandas, or similar computational tools.
• Solid understanding of algorithms, numerical methods, and computational problem-solving techniques.
• Demonstrated ability to design complex, original mathematical problems with clear and unique solutions.
• Excellent technical writing skills with strong attention to clarity, precision, and correctness in English.
• Experience with mathematical modeling, optimization, symbolic computation, or related research tools is highly desirable.
• Familiarity with olympiad-level, graduate-level, or research-level problem design is a strong advantage.
• Prior experience in AI evaluation, benchmarking, or dataset creation is a plus.

Benefits:

• Competitive hourly compensation up to 40 USD per hour
• Flexible remote contract arrangement with approximately 20 hours per week commitment
• Short-term project (2 months) with potential for extension based on performance and demand
• Opportunity to contribute directly to frontier AI model development and evaluation
• High-impact research work applied to real-world AI reasoning systems
• Exposure to advanced AI training pipelines and cutting-edge mathematical datasets
• Flexible working schedule with fully remote setup