About this opportunity
Description
About the Role
Turing is looking for Web Research Specialists to help build evaluation benchmarks for advanced AI browsing agents.
This is an investigative research role, not a traditional content-writing or subject-matter expert position. You'll design challenging research problems that test whether frontier AI systems can locate and reason through difficult-to-find information on the open web.
You'll begin with a verifiable fact, work backwards to construct a difficult research question, and provide a complete evidence trail demonstrating how the answer can be independently verified.
What You'll Do
- Create natural-language research questions with short, stable, objectively verifiable answers.
- Develop independently verifiable clues involving dates, people, places, organisations, events, records, quantities, and other fact types.
- Conduct extensive open-web research across multiple sources.
- Navigate government and institutional databases, archives, registries, PDFs, and primary records.
- Document the searches performed and the results obtained.
- Provide precise citations to specific pages, tables, sections, and records.
- Build clear and auditable evidence trails supporting each research problem.
Requirements
- Master's degree OR more than 3 years of relevant experience.
- Strong open-web and investigative research skills.
- Ability to locate and work with primary records and difficult-to-find sources.
- Native or near-native written English.
- Strong attention to detail and sourcing accuracy.
- Comfortable researching unfamiliar subjects from scratch.
- Comfortable producing detailed, structured documentation.
- Experience with LLM evaluation, red-teaming, or benchmark construction.
- Relevant experience in areas such as investigative journalism, fact-checking, OSINT, archival research, legal/patent research, genealogy, due diligence, competitive quizzing, or puzzle construction.
Familiarity with JSON and structured data formats is a plus.
Contract Details
- 40 hours per week
- At least 4 hours of PST overlap
- 8-week contract
- Contractor engagement
- No medical or paid leave
Application Process
Applicants will complete an assessment. Turing will evaluate the submission and contact selected candidates regarding the next steps.
