AI Trainer – Computer Science Expert - Graphical Abstract - Fully Remote - (AUS)

Role overview

A global human data platform is looking for computer science specialists based in Australia to act as expert evaluators of AI-generated graphical abstracts. The role involves checking whether machine-produced visual summaries of research papers are scientifically accurate and structurally sound. It is fully remote, flexible, and project-based, with expert input directly influencing the quality of future AI summarisation.

Responsibilities

- Read peer-reviewed computer science papers and study their corresponding AI-generated graphical abstracts. - Verify that the abstract's text and visuals align with the paper's claims, methodology, and conclusions. - Detect factual mistakes, omissions, or distortions introduced by the model. - Assess the layout and information hierarchy of each abstract for clarity and faithfulness to the source. - Comment on whether domain-specific terminology is used appropriately throughout the summary. - Submit structured evaluations that help train and validate AI systems.

Requirements

- A Master's or PhD in Computer Science or a closely related field. - Deep familiarity with the conventions of computer science research writing. - A track record of analytical work such as fact-checking, reviewing, or technical editing. - Strong written communication skills for providing clear, concise feedback. - Self-directed work habits suitable for remote, asynchronous collaboration. - Residency in Australia and the right to work as an independent contractor on a per-task basis.

Nice to have

- Previous experience annotating or evaluating data for AI/ML training pipelines. - Background in scientific publishing, conference reviewing, or editorial work.

Benefits and work setup

- Fully remote with no fixed hours; tasks are assigned on an ad hoc basis. - Competitive compensation paid per completed task. - Streamlined onboarding, generally completed shortly after passing the qualification assessment.