Web Researcher – AI Benchmarking

This is a fully remote position, open to applicants in Poland.

📋 Description

• Create complex open-web research challenges to assess advanced AI browsing agents.

• Begin with objectively verifiable information and formulate questions that make those facts hard to uncover.

• Develop multi-layered clue frameworks that encompass dates, individuals, locations, organizations, works, events, records, and quantities.

• Ensure that each clue can be independently verified and adheres to explicit constraints.

• Conduct research utilizing government databases, institutional resources, archives, registries, and PDF files.

• Generate comprehensive evidence trails that reference precise pages, tables, sections, or records.

• Document validation searches and note the outcomes of standard search methods.

• Evaluate if research questions remain challenging through multiple search attempts.

• Refine questions to remove ambiguity while maintaining their level of difficulty.

• Provide organized, reproducible research documentation.

⛳️ Requirements

• A background in reference librarianship, archives, or special collections.

• Experience in investigative journalism or professional fact-checking.

• Familiarity with OSINT, due diligence, KYC, or investigative research.

• A background in patent, prior-art, or legal-discovery research.

• Experience with genealogy or historical records research.

• Involvement in competitive quizzing, puzzle design, or puzzle-hunt creation.

• Knowledge of JSON or structured data formats.

• Availability of 40 hours per week, with a minimum of 4 hours per day, including 4 hours of overlap with PST.

🏝️ Benefits

• Opportunity to work on cutting-edge AI research.

• Engage in intellectually stimulating and challenging projects.

• Collaborate with a diverse team of experts in the field.