Apply GPU kernel development and parallel computing expertise remotely from Canada as a freelance AI Task Auditor. Evaluate technical tasks for accuracy and reliability.

This partner-managed freelance project suits experienced technical professionals with GPU kernel development experience. You will assess AI training tasks, examine realistic engineering challenges, and provide evidence-based feedback. Your work supports stronger evaluation benchmarks and more reliable AI training workflows.

Key Responsibilities:

- Evaluate GPU kernel tasks for accuracy, realism, and reproducibility
- Verify tests, evaluation criteria, and expected outcomes
- Identify memory bugs, bottlenecks, and parallel logic errors
- Troubleshoot scenarios and assess proposed technical solutions




- Provide actionable feedback on task quality and reliability

Requirements:

- Professional GPU kernel development or related experience
- Knowledge of CUDA, Triton, parallel computing, or optimization
- Strong analytical, critical-thinking, and problem-solving skills
- Ability to test and troubleshoot complex technical scenarios
- Secure computer and reliable high-speed internet connection

Use GPU computing expertise, systematic analysis, and accurate communication to improve AI training and evaluation workflows remotely.

📌 GPU Kernel AI Task Auditor (Montreal)
🏢 Jobgether
📍 Montreal