AI Training Specialist – Chemistry & Python | Contractor | Worldwide Remote
Turing, the world's leading research accelerator for frontier AI labs, is seeking highly specialized Subject Matter Experts (SMEs) in Chemistry (and related STEM fields) to contribute to a cutting-edge scientific benchmark development project. This is a high-impact role where you will design research-grade computational science problems used to evaluate and train advanced AI systems, including Gemini 3 Pro.
About Turing
Based in San Francisco, CA, Turing partners with frontier AI labs and global enterprises to accelerate AI research and deployment. Our experts specialize in coding, reasoning, STEM, and more — helping transform AI from proof of concept into measurable business intelligence.
Role Overview
You will design benchmark-quality computational science problems modeled after the SciCode framework — a research-grade benchmark for evaluating AI on real computational science challenges. Each problem must include a detailed prompt, a verified Python reference solution, and a rigorous automated test suite.
Key Responsibilities
- Select a scientific domain and subdomain (e.g., Reaction Kinetics, Quantum Chemistry, Computational Genomics)
- Design computationally intensive main problems with clear problem statements, scientific background, function signatures, and all necessary parameters
- Implement complete, correct Python reference solutions using NumPy, SciPy, and related libraries
- Write comprehensive pytest test suites validating correctness, numerical accuracy, and algorithmic behavior
- Create 2–4 structured subproblems per main problem, building scaffolded evaluation pipelines
- Deliver each problem set as a well-organized Jupyter Notebook (.ipynb) with Prompt, Background, Solution, and Testing cells
- Ensure diversity across topics — no repetition of similar problems or algorithmic patterns
Problem Quality Standards
- Computationally Intensive: Must require non-trivial implementations (iterative solvers, numerical simulations, matrix decompositions, adaptive methods)
- Python Code Output: Solutions must be standalone functions using scientific libraries
- Verifiable Results: Problems must have precise, testable numerical outcomes
Required Qualifications
- PhD in Chemistry, Physics, Mathematics, Biology, or a related computational field (preferred)
- Master's degree with 5+ years of research or industry experience in computational science may be considered
- Strong proficiency in Python for scientific computing
- Experience with NumPy, SciPy, SymPy, and Pandas
- Deep understanding of numerical methods, algorithms, and computational complexity
- Experience writing automated tests with pytest
- Familiarity with numerical stability, floating-point arithmetic, and error analysis
This is a contractor opportunity open to qualified experts worldwide. If you have a strong background in computational chemistry or a related STEM field and want to shape the future of AI evaluation, we encourage you to apply.
