Remote Pod Lead - Math

Turing, a rapidly growing leader in AI innovation, is advancing the development and deployment of cutting-edge AI systems. The company collaborates with top AI labs to push the boundaries of model capabilities in areas such as reasoning, coding, multilinguality, and STEM knowledge. These advancements are then applied to create real-world AI solutions that address critical business challenges.

Key Responsibilities:
Lead a team of 5–10 trainers focused on developing realistic, terminal-based scientific and technical tasks for Terminal Bench Science. Review trainer-created tasks thoroughly to ensure mathematical accuracy, reproducibility, challenge level, and grader reliability. Mentor trainers, manage task quality and throughput, and ensure tasks reflect authentic mathematical research and computational workflows. Oversee tasks spanning numerical and applied mathematics, scientific computing, optimization, statistics, mathematical modeling, computational algebra, and formal verification.

Responsibilities include reviewing all aspects of tasks, such as problem statements, datasets, environments, reference solutions, and automated tests. Verify mathematical correctness, well-posedness, convergence, stability, conditioning, discretization error, and edge-case handling. Ensure tasks require genuine multi-step reasoning rather than simple lookups or single library calls. Validate the appropriate use of exact, symbolic, arbitrary-precision, or floating-point computations and confirm that tolerances align with the problem’s numerical conditioning.

Design and review graders based on mathematical properties like residuals, feasibility, optimality certificates, invariants, error bounds, or proof checking. Guarantee reproducibility by using pinned tools, libraries, solver versions, and dependencies. Detect and prevent hard-coded solutions, test-file leakage, and other methods that bypass intended computations. Provide precise, actionable feedback and track trainer revisions until closure.

Mentor and onboard trainers on mathematical rigor, task design, coding practices, and evaluation standards. Allocate work, monitor quality and throughput, escalate blockers, and maintain pod documentation covering guidelines, recurring issues, and exemplary tasks. Collaborate closely with the Mathematics Team Lead and other technical stakeholders to continuously improve benchmark quality.

Required Qualifications:
Ph.D., postdoctoral experience, or equivalent advanced experience in mathematics, applied mathematics, statistics, or a related discipline. Strong programming skills in Python, Julia, C/C++, MATLAB/Octave, or computer algebra systems. Proficiency with Linux and terminal environments. Hands-on experience with numerical methods, scientific computing, symbolic computation, optimization, computational statistics, or mathematical modeling. Ability to rigorously evaluate mathematical derivations, algorithms, proofs, and computational results.

Experience reviewing research outputs, code, mathematical work, or scientific workflows is essential. Proven ability to mentor or lead small technical or research teams. Strong written communication skills with the ability to provide precise, constructive feedback. Capability to assess reproducibility, test reliability, and the scientific validity of computational tasks.

Preferred Qualifications and Benefits:
Experience with scientific computing libraries and tools such as NumPy, SciPy, SymPy, mpmath, SageMath, PARI/GP, GAP, Macaulay2, or Mathematica-like workflows is advantageous. Familiarity with ODE/PDE tools like FEniCS, Firedrake, DifferentialEquations.jl, or PETSc is a plus. Knowledge of optimization frameworks such as CVXPY, JuMP, Pyomo, or HiGHS is beneficial. Experience with proof assistants like Lean/mathlib, Coq/Rocq, or Isabelle is desirable.

Additional experience with Docker, Git, continuous integration systems, and automated testing will be valuable. Background in designing benchmarks, competition problems, or computational evaluations is preferred. Experience evaluating AI coding, reasoning, or terminal agents, as well as setting or judging mathematics olympiad or competition problems, is a strong plus. Contributions to publications or open-source projects in computational mathematics, numerical software, or formalized mathematics are welcomed. Prior involvement in AI training, data generation, or model evaluation programs is also beneficial.

Offer Details:
This is a contractor assignment requiring a commitment of at least 4 hours per day and a minimum of 40 hours per week, with a 4-hour overlap with Pacific Standard Time (PST). The contract duration is 4 weeks, with an expected start date as early as next week. Please note that this role does not include medical benefits or paid leave.