AI Trainer – Computer Science Expert - Graphical Abstract - Fully Remote Boston

A human data platform is recruiting computer science specialists in the Boston area to evaluate graphical abstracts produced by an AI system. The position focuses on verifying scientific accuracy, completeness, and visual coherence of machine-generated summaries of research papers. It is a flexible, fully remote, ad hoc opportunity aimed at academics and industry researchers who want to contribute expert feedback to AI development.

• Read computer science research papers and compare them against their AI-generated graphical abstracts.
• Determine whether each abstract faithfully represents the paper's main contributions and methodology.
• Identify errors, omissions, or misleading framings introduced by the model.
• Assess the abstract's layout, sequencing, and balance of text versus visuals.
• Verify that specialised terminology is used precisely and in the correct context.
• Document evaluations clearly so they can be used to improve future model outputs.

• A Master's or PhD in Computer Science or a closely related field.
• Demonstrated ability to interpret and critique peer-reviewed technical research.
• Experience with fact-checking, technical reviewing, or similar detail-oriented analytical work.
• Excellent written communication and consistent attention to nuance.
• Comfort working autonomously in a remote, asynchronous environment.
• Located in the Boston area, with eligibility to work on an independent contractor basis.

• Prior exposure to AI training data annotation, RLHF, or model evaluation.
• Background in scientific publishing, conference reviewing, or journal editing.

• Fully remote with ad hoc task allocation and flexible scheduling.
• Competitive pay rates per completed task.
• Fast onboarding, typically finalised shortly after a brief qualification assessment.