STEM Chemistry – CUDA/Python AI Trainer at Turing

CUDA/Python AI Trainer | Contractor | Remote (Select Countries)

Join Turing — one of the world's fastest-growing AI companies — as a freelance STEM specialist helping leading AI labs improve their Large Language Models. In this role, you'll translate CUDA/C++ code into Python using PyTorch and NumPy, evaluate LLM-generated code, and contribute to RLHF workflows that shape the future of frontier AI systems.

About Turing

Turing accelerates the advancement and deployment of powerful AI systems by partnering with the world's top AI labs to improve model reasoning, coding, multimodality, and STEM capabilities — and by building real-world AI solutions for enterprise clients.

Role Overview

This is a data generation role supporting foundational LLM companies. You will produce high-quality proprietary data used for fine-tuning and benchmarking AI models. Work spans both SFT (Supervised Fine-Tuning) and RLHF (Reinforcement Learning from Human Feedback) workflows. Note: This role does not involve building or fine-tuning LLMs directly.

Day-to-Day Responsibilities

• Translate CUDA/C++ code into equivalent Python implementations using PyTorch and NumPy, ensuring logical and performance parity.
• Analyze CUDA kernels and GPU-accelerated code for structure, efficiency, and function prior to translation.
• Evaluate LLM-generated CUDA/C++ to Python translations, providing technical feedback and corrections.
• Collaborate with prompt engineers and researchers to develop real-world CUDA/PyTorch test prompts.
• Participate in RLHF workflows — rank LLM responses and justify ranking decisions clearly.
• Debug and review translated Python code for correctness, readability, and industry-standard consistency.
• Maintain technical documentation to support reproducibility and code clarity.
• Propose improvements to prompt structure or conversion approaches based on observed LLM failure patterns.

Requirements

• 5+ years of overall work experience, including at least 3 years in Python and 2+ years in CUDA/C++.
• Strong hands-on experience with scientific computing using PyTorch and NumPy.
• Solid understanding of CUDA programming concepts and C++ fundamentals.
• Demonstrated ability to analyze CUDA kernels and accurately reproduce them in Python.
• Familiarity with GPU computation, parallelism, and performance-aware coding practices.
• Strong debugging skills with attention to numerical consistency when porting logic across languages.
• Experience evaluating AI-generated code or participating in LLM tuning is a plus.
• Fluent in written and conversational English.

Engagement Details

• Commitment: Minimum 20 hrs/week (options: 20, 30, or 40 hrs/week); at least 4 hrs/day with 4-hour PST overlap required.
• Type: Contractor assignment (no medical/paid leave benefits).
• Duration: 3-month contract, with an expected start date of next week.
• Eligible Locations: India, Pakistan, Nigeria, Kenya, Egypt, Ghana, Bangladesh, Turkey, Brazil, Mexico.

Evaluation Process (~75 minutes)

• Round 1: 60-minute technical interview
• Round 2: 15-minute cultural fit and offer discussion

Perks

• Fully remote, flexible work environment.
• Opportunity to contribute to cutting-edge AI projects with leading LLM companies.