Remote Pod Lead - Math

Turing is a rapidly growing AI company focused on advancing and deploying powerful AI systems. The company collaborates with leading AI labs to push the boundaries of model capabilities in areas such as reasoning, coding, multimodality, and STEM knowledge. These advancements are then applied to develop real-world AI solutions that address critical business challenges.

Key Responsibilities:
Lead a team of 5–10 trainers in creating realistic, terminal-based scientific and technical tasks for Terminal Bench Science. Review tasks thoroughly to ensure mathematical accuracy, reproducibility, challenge level, and grader reliability. Mentor trainers on mathematical rigor, task design, coding standards, and evaluation criteria. Manage quality control and task throughput while maintaining detailed documentation on guidelines and recurring issues. Collaborate with the Mathematics Team Lead and other stakeholders to enhance benchmark quality. Evaluate all aspects of tasks including problem statements, datasets, environments, reference solutions, and automated tests. Verify mathematical properties such as convergence, stability, conditioning, and error handling. Ensure tasks require genuine multi-step reasoning and appropriate use of exact, symbolic, or floating-point computations. Design and review graders based on mathematical principles like residuals, error bounds, and proof checking. Confirm reproducibility by controlling tool versions and dependencies. Detect and prevent shortcuts such as hard-coded answers or test-file leakage. Provide precise, actionable feedback and track revisions to closure. Allocate work effectively, monitor team performance, and escalate blockers as needed.

Required Qualifications:
Ph.D., postdoctoral experience, or equivalent advanced expertise in mathematics, applied mathematics, statistics, or a related discipline. Proficiency in programming languages such as Python, Julia, C/C++, MATLAB/Octave, or computer algebra systems. Strong command of Linux and terminal environments. Practical experience with numerical methods, scientific computing, symbolic computation, optimization, computational statistics, or mathematical modeling. Ability to critically assess mathematical derivations, algorithms, proofs, and computational results. Experience reviewing research outputs, code, or scientific workflows. Demonstrated leadership or mentoring experience in small technical or research teams. Excellent written communication skills with the ability to deliver clear, constructive feedback. Competence in evaluating reproducibility, test reliability, and scientific validity of computational tasks.

Preferred Qualifications and Benefits:
Familiarity with libraries and tools such as NumPy, SciPy, SymPy, mpmath, SageMath, PARI/GP, GAP, Macaulay2, or Mathematica-like environments. Experience with ODE/PDE frameworks like FEniCS, Firedrake, DifferentialEquations.jl, or PETSc. Knowledge of optimization frameworks including CVXPY, JuMP, Pyomo, or HiGHS. Exposure to formal proof assistants such as Lean/mathlib, Coq, Isabelle, or Rocq. Experience with Docker, Git, continuous integration systems, and automated testing. Background in designing benchmarks, competition problems, or computational evaluations. Prior involvement in evaluating AI coding, reasoning, or terminal agents. Participation in mathematics olympiad problem setting or judging. Publications or open-source contributions in computational mathematics, numerical software, or formalized mathematics. Previous experience in AI training, data generation, or model evaluation programs is a plus.

Offer Details:
This is a contractor role requiring a minimum commitment of 40 hours per week, with at least 4 hours of daily overlap with Pacific Standard Time (PST). The contract duration is 4 weeks, with an expected start date in the near term. Please note this is a contractor assignment without medical benefits or paid leave.