Remote SWE Agent Workflow Reviewer - Code & Test Quality

AI Trainer Jobs; United States; Remote; USD 152,000 - 207,000; Full time

Job summary

AI Trainer Jobs is seeking a SWE Agent Workflow Reviewer for a remote independent contractor role. You will reproduce failures, write the unit tests the model should have written, and explain fixes so the modeling team can target the gap.

Engineering model quality hinges on code compiling, tests passing, and edge cases handled in sandboxed environments. AuraOne pairs experienced engineers with the modeling team to grade outputs the way a code reviewer would, judging generated code and software

Qualifications

• Strong engineering experience to read, run, and debug unfamiliar code.
• Ability to write concise unit tests capturing a single failure mode.
• Experience with a testing framework such as pytest, Jest, JUnit, Go test, or RSpec.
• Clear written reasoning to convince a senior engineer.
• Reliable async availability for at least 10 hours per week.
• Prior code-review or technical-interview-grading experience is a plus.

Responsibilities

• Run and reproduce candidate code outputs in a sandboxed environment.
• Grade SWE agent workflow solutions for correctness, style, and edge-case handling.
• Write the smallest failing tests that demonstrate a bug.
• Compare paired solutions and rank them with a written rubric-based rationale.
• Triage security issues and document the patch the model should have produced.

Skills

Code review

Debugging

Unit testing

Software testing

Async availability

Open-source contributions

Clear written reasoning

Tools

pytest

Jest

JUnit

Go test

RSpec

Job description

AI Trainer Jobs is seeking a SWE Agent Workflow Reviewer for a remote independent contractor role. You will reproduce failures, write the unit tests the model should have written, and explain fixes so the modeling team can target the gap.

Engineering model quality hinges on code compiling, tests passing, and edge cases handled in sandboxed environments. AuraOne pairs experienced engineers with the modeling team to grade outputs the way a code reviewer would, judging generated code and software