Remote SWE-Bench AI Task Auditor - Freelance

Meridial; Austin (TX); On-site; USD 68,000 - 97,000; Full time

Job summary

Meridial is seeking a SWE-Bench Task Auditor for a freelance AI trainer project. The role involves auditing tasks used to train AI systems, ensuring technical rigor, practicality, and accuracy across specialized software engineering domains.

You will provide clear feedback on codebase integration issues, test failures, or logic errors, and work remotely as a contractor. The pay rate is $60 per hour, with final terms determined after evaluating experience and location.

Qualifications

• Proven experience in software engineering with complex codebase navigation.
• Real-world application development experience.
• Familiarity with SWE-Bench standards.

Responsibilities

• Audit SWE-Bench tasks for technical accuracy, realism, solvability, reproducibility, and test support.
• Provide actionable feedback on codebase integration issues, test failures, or logic errors.

Skills

Software Engineering

Codebase Navigation

Real-world App Development

SWE-Bench Standards

Analytical Skills

Troubleshooting

Job description

Meridial is seeking a SWE-Bench Task Auditor for a freelance AI trainer project. The role involves auditing tasks used to train AI systems, ensuring technical rigor, practicality, and accuracy across specialized software engineering domains.

You will provide clear feedback on codebase integration issues, test failures, or logic errors, and work remotely as a contractor. The pay rate is $60 per hour, with final terms determined after evaluating experience and location.