QA Engineer

What you'll do

The brief that matters is in one line of the ad: identify answers that may appear correct but are actually inaccurate or incomplete. That is the QA instinct pointed at model output, where the plausible pass is the dangerous one, and it is the reason this project wants testers rather than general reviewers.

Around that sits work any QA engineer will recognise. You rate AI-generated technical outputs against defined criteria and rubrics. You design and assess test cases across functional, regression and edge-case scenarios, including negative and boundary tests. You review bug reports and test documentation to check that issues are actually reproducible, that the documentation is complete and that severity has been judged correctly. You isolate and document defects with precise reproduction steps in structured tracking, and write feedback specific enough that a developer can act on it without coming back to ask what you meant.

You also help refine the evaluation guidelines themselves, so the standards improve rather than just get applied.

What you need

• Proven professional experience as a QA Engineer, SDET, Test Engineer, QA Analyst or similar
• A strong command of QA fundamentals: test case design, bug tracking, regression testing, and both manual and automated strategies
• Hands-on experience with automation frameworks and test management tools. The ad names Selenium, Playwright, Cypress, Appium, Postman, Jira, TestRail, Zephyr and BrowserStack, or similar
• Excellent analytical and problem-solving skills with meticulous attention to detail
• Written English at B2 level or above, giving specific and actionable feedback
• A reliable internet connection and readiness to begin promptly

Prior paid human data experience supporting AI training (annotation, labelling, RLHF, AI response evaluation, model evaluation or rubric-based grading) is a plus rather than a requirement.

No formal degree is required. The ad states that practical, demonstrable testing experience takes precedence, which is more explicit than most listings get and worth taking at face value.

What it pays

$90–175/hour, micro1's own published figure. That is strong against QA contract rates in most markets, and the ceiling is the highest on any of micro1's current evaluation listings.

The band is $85 wide, close to double from floor to ceiling, with nothing published about what decides placement. Automation depth and prior paid AI-training work are the plausible levers, but the ad does not say so, so plan around $90 until you have an offer. No compensation structure is stated, so whether the rate buys your time or works out per task is unanswered.

Worth knowing

Good:

• $90–175/hour is well above typical QA contracting in most markets
• Thirty openings
• No formal degree required, stated explicitly, with demonstrable testing experience taking precedence
• Location is given only as remote, with no country list attached
• B2 written English rather than native-level fluency, a noticeably lower language bar than micro1's writing-evaluation roles
• Prior AI-training experience is a plus, not a gate, so this is an entry point into the field on an existing skill
• The work is recognisable QA practice, not a new discipline to learn

Less good:

• An $85-wide band with no published basis for placement; the ceiling is nearly double the floor
• You are evaluating and documenting, not testing a product you can influence
• No compensation structure, project duration, weekly hours or start date stated
• "Readiness to begin promptly" implies an immediate start rather than a planned one
• High-volume rating work is repetitive by design
• Contractor engagement: no benefits, no notice period, your own tax to manage
• The customer project is not named

About this listing

Posted by micro1 on its own jobs portal, confirmed open on 12 September 2026. The $90–175/hour band, the thirty openings, the tool list, the B2 language requirement and the explicit absence of a degree requirement are all its own published text for this listing.

Location is given only as remote with no country list, so eligibility is left open rather than filled with a guess. No pay structure, project duration or weekly hours are published, and the customer is not named. Screening and payment run at the platform level. See micro1.