The work
iMerit, an EXL company, is looking for detail oriented analysts to evaluate and rank AI generated responses to image based prompts. You will judge answers on accuracy, relevance, clarity, conciseness, safety, localization, and how well they follow the user's instructions, then explain your reasoning in writing.
Much of the job comes down to this: look at the image, look at what the model said about it, and decide whether the two actually match.
What you will do
Review images alongside the prompts and responses attached to them
Interpret both visual and conversational context before scoring anything
Compare multiple responses and rank them against defined criteria
Confirm that descriptions and claims about an image reflect what is actually there
Verify factual claims using approved research sources
Flag tasks that cannot be reliably evaluated
Write clear justification for every ranking you submit
What you bring
A sharp eye for visual detail and the patience to look twice
Strong critical thinking, research skills, and reading comprehension
Sound judgment and well supported decision making
Professional written English, including the ability to describe what you see precisely
The consistency to apply detailed guidelines the same way, task after task
Experience in image annotation, content review, editing, research, design, or localization is welcome but not essential.
Good to know
Independent contractor engagement for the length of the project.
Fully remote with flexible scheduling around agreed deadlines.
Tasks include visual content, so a device and connection that display images comfortably are required.
Volume varies with project demand.
The rate of pay will be specified in the job offer
iMerit Technology, an EXL company, specializes in AI training and evaluation services. The company employs analysts to evaluate and rank AI-generated responses to image-based prompts, assessing accuracy, relevance, clarity, safety, and instruction-following. Their work involves comparing multiple AI responses against defined criteria and verifying factual claims about visual content using approved research sources. iMerit focuses on improving AI model performance through detailed human evaluation and feedback on visual understanding and conversational accuracy.
