AI Evaluation Specialist (CS) — Remote

Prolific Academic Ltd; United States; Remote; USD 68,000 - 97,000; Full time

Benefits offered by this job

Competitive pay

Flexible hours

Work from home

Job summary

Prolific is seeking Computer Science Specialists to join our Expert Network to help train and evaluate AI models. You’ll complete a short test and, if successful, be invited to participate as paid contributor with flexible hours and the ability to work from home.

Researchers looking for your skills tend to pay up to $60/hr, depending on skills and experience level. You must be prepared to commit one hour of uninterrupted work for paid tasks, with many tasks shorter.

Qualifications

• BSc (Bachelor of Science) in Computer Science or related field.
• Ability to interpret research papers, understand complex algorithms, and review code logic.
• Analytical mindset with attention to technical gaps and logical flaws.
• Valid LinkedIn profile to verify degree and background during screening.
• PayPal account to receive payment.

Responsibilities

• AI Evaluation & Ranking: compare AI responses and rank by accuracy, logic, and safety.
• Scientific Review: assess CS papers and AI summaries for integrity.
• Fact-Checking: identify inaccuracies in data, formulas, or findings.
• RLHF: provide human ground truth to align models with software standards.
• Code & Logic Verification: audit AI-generated code and diagrams for correctness.

Skills

CS concepts

Research interpretation

Algorithm understanding

Code review

Education

BSc in Computer Science

Job description

Prolific is seeking Computer Science Specialists to join our Expert Network to help train and evaluate AI models. You’ll complete a short test and, if successful, be invited to participate as paid contributor with flexible hours and the ability to work from home.

Researchers looking for your skills tend to pay up to $60/hr, depending on skills and experience level. You must be prepared to commit one hour of uninterrupted work for paid tasks, with many tasks shorter.