GPU Kernels AI Task Auditor - Freelance AI Trainer Project

Meridial; United States; Remote; USD 96,000 - 138,000; Part time

Job summary

Meridial is seeking experienced GPU kernel specialists to audit AI training tasks for technical rigor, realism, solvability, and reproducibility. You will assess task accuracy and provide actionable feedback on memory issues, bottlenecks, and parallel logic.

The freelance contract role is remote in the United States, with hourly pay in a $70–$100 range; final offers reflect experience and geography. Contractors supply their own equipment; standard company health benefits do not apply.

Qualifications

• Demonstrable professional experience and deep technical knowledge in GPU Kernel development (including CUDA/Triton programming, parallel computing, and hardware-level performance optimization).
• Strong analytical and problem-solving skills to rigorously test, troubleshoot, and evaluate complex technical scenarios.

Responsibilities

• Task Evaluation: Assess whether GPU kernel tasks are technically accurate, realistic, solvable, reproducible, and supported by reliable tests and evaluation criteria.
• Technical Auditing & Feedback: Provide clear, actionable feedback on memory allocation bugs, execution bottlenecks, or parallel computing logic errors within the tasks.

Skills

GPU Kernel development

CUDA/Triton programming

Parallel computing

Hardware-level performance

Tools

CUDA

Triton

Job description

Project Overview We are sourcing experienced technical specialists in GPU Kernels to audit tasks used to train and evaluate AI systems. The objective of this project is to ensure that AI training workflows and tasks are technically rigorous, practical, and highly accurate within your specific area of expertise.

Project Deliverables & Scope

• Task Evaluation: Assess whether GPU kernel tasks are technically accurate, realistic, solvable, reproducible, and supported by reliable tests and evaluation criteria.
• Technical Auditing & Feedback: Provide clear, actionable feedback on any identified memory allocation bugs, execution bottlenecks, or parallel computing logic errors within the tasks.

Required Expertise

• Demonstrable professional experience and deep technical knowledge in GPU Kernel development (including CUDA/Triton programming, parallel computing, and hardware-level performance optimization). (Note: Candidates will be considered for one specialty based on their experience; expertise across other domains is not required).
• Strong analytical and problem-solving skills to rigorously test, troubleshoot, and evaluate complex technical scenarios.

We offer a pay range of $70 - $100per hour, with the exact rate determined after evaluating your experience, expertise, and geographic location. Final offer amounts may vary from the pay range listed above. As a contractor you’ll supply a secure computer and high‑speed internet; company‑sponsored benefits such as health insuranceand PTO do not apply.

Employment type: Freelance Contract

Workplace type: Remote