AI Benchmark & Red-Teaming Specialist

Location: Remote, US; requires up to 4 hours of overlap with PST.

Company

Turing accelerates frontier AI research and helps enterprises deploy advanced AI systems through high-quality data, training pipelines, and AI research expertise.

What you will do

• Research niche topics and identify factual ground truths that are difficult to find through direct searches.
• Verify facts against primary or highly credible sources, cross-check contradictory evidence, and document source URLs.
• Design clear, answerable benchmark tasks that expose structural weaknesses in AI research strategies.
• Run tasks against designated state-of-the-art models and confirm intended failures.
• Inspect reasoning chains, search logs, trajectories, and intermediate outputs to diagnose model failures.
• Apply validated model-breaking patterns across domains and review peer-authored tasks when performing strongly.

Requirements

• At least 1 year of experience in investigative research, fact-checking, AI benchmark task authoring, red-teaming, or technical QA review.
• Proven experience authoring, testing, or reviewing tasks for AI benchmarks such as GAIA, BrowseComp, SimpleQA, SWE-bench, Terminal-Bench, or Humanity’s Last Exam.
• Intermediate-to-advanced web search and Open Source Intelligence (OSINT) skills, including navigating academic archives, local records, or obscure databases.
• Strong analytical ability to inspect model reasoning chains, search logs, and source selection and diagnose root causes of failure.
• Ability to write clear, unambiguous benchmark prompts with primary-sourced answer keys.
• A bachelor’s degree or higher in a relevant engineering domain is required.

Nice to have

• Experience designing prompt-injection or reasoning-trap tasks that break state-of-the-art models.
• Advanced OSINT capabilities or published investigative or academic research demonstrating deep source verification.
• Niche expertise in hard-to-find records, such as local government filings, regional sports history, legal archives, or historical patents.
• Experience as a lead reviewer, benchmark auditor, or quality owner on AI evaluation projects.

Culture & Benefits

• Contractor engagement for up to 4 weeks.
• Commitment of up to 40 hours per week.
• Remote work in the US with 4 hours of PST overlap.

Hiring process

• Submit a Job Interest Form and complete an assessment.
• Candidates finalized through the assessment proceed to review and subsequent stages.