About The Position

Weekday is recruiting for a client seeking PhD-level subject matter experts for a paid machine learning model evaluation project. Experts will assess the quality, accuracy, and reliability of AI-generated responses to research-focused prompts and domain-specific questions, helping identify where models perform effectively and where their outputs need refinement. The role is suited to experienced researchers, academics, data analysts, technical writers, and domain specialists who can critically assess complex information, identify factual inaccuracies, and provide structured, evidence-based feedback. The engagement is an independent contractor project expected to involve approximately five hours of evaluation work. It is fully remote, flexible, and asynchronous. Payment is $150 per hour and is processed weekly based on services rendered. The engagement may be extended, shortened, or concluded early depending on project requirements and individual performance. The posting notes that candidates should be able to start by Friday at end of day, but does not specify a calendar date. Applicants must be based in the US, UK, Canada, the EU, Norway, Switzerland, or Iceland, and should have a finished or ongoing paper in a topic associated with NeurIPS, ICML, ICLR, COLM, or RLC. H-1B and STEM OPT candidates cannot be supported. Participation does not require access to confidential or proprietary information belonging to a current or former employer, client, or academic institution. During interviews, candidates may be assessed on their review practices, critical thinking, and ability to improve evaluation quality. The posting identifies three habits for discussion: exchanging unfinished review notes with another reviewer, maintaining a separate log of opinions or assessments that change during the review process, and using review notes and feedback to update personal reading lists.

Requirements

  • PhD or equivalent advanced subject matter expertise in a relevant academic, technical, or professional discipline.
  • Strong academic or professional background in a specialised domain and ability to critically assess complex information.
  • Excellent analytical, research, and critical-thinking skills.
  • Ability to evaluate technical content, research findings, and domain-specific information for factual correctness and relevance.
  • Strong written communication skills, particularly for clear and structured evaluation justifications.
  • High attention to detail and ability to identify subtle inaccuracies, inconsistencies, and gaps in AI-generated responses.
  • Ability to work independently and complete approximately five hours of asynchronous evaluation tasks.
  • Applicants must be based in the US, UK, Canada, the EU, Norway, Switzerland, or Iceland.
  • Should have a finished or ongoing paper in a topic associated with NeurIPS, ICML, ICLR, COLM, or RLC.
  • H-1B and STEM OPT candidates cannot be supported.

Nice To Haves

  • Background as an academic researcher, university faculty member, PhD scholar, doctoral-level subject matter expert, data analyst, research analyst, technical writer, research professional, or domain specialist experienced in evaluating technical or research-focused content.
  • Ability to start on the project by Friday at end of day.
  • Review practices to discuss during interviews include exchanging unfinished review notes with another reviewer, keeping a separate log of assessments that change, and using review notes and feedback to update personal reading lists.

Responsibilities

  • Review machine learning model outputs generated in response to research-focused and domain-specific prompts.
  • Evaluate AI-generated responses for factual accuracy, relevance, completeness, and overall quality.
  • Assess the technical validity and reliability of information in model-generated responses.
  • Use a structured rating system to evaluate and score responses consistently.
  • Provide concise, clear, well-reasoned written justifications for evaluation scores.
  • Identify factual errors, gaps in reasoning, irrelevant information, and areas requiring improvement.
  • Complete approximately five hours of independent evaluation work asynchronously.
  • Maintain consistency and attention to detail throughout the evaluation process.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service