Web Research Specialist

About the Company

Turing is among the fastest expanding AI firms worldwide, focused on accelerating the development and implementation of advanced AI systems. They partner with leading AI research labs to enhance emerging AI capabilities such as reasoning, coding, multilingualism, and STEM knowledge, and apply these advancements to build practical AI solutions addressing critical business needs.

Role Overview

The role involves creating evaluation benchmarks for advanced AI web browsing agents. The specialist will design challenging research queries that state-of-the-art AI systems cannot easily resolve, even when given unrestricted web access and multiple search attempts. This position emphasizes investigative research rather than content expertise or writing.

Key Deliverables

• Develop natural language research questions with clear, concise, and objectively verifiable answers.
• Provide multiple independently verifiable clues covering facts such as dates, people, locations, organizations, works, events, records, and quantities, each with specific constraints.
• Compile validation records that document the initial searches performed and their outcomes.

Required Qualifications

• Demonstrated skill in open-web research involving locating primary documents and navigating institutional and government databases, archives, registries, and PDFs.
• Meticulous citation practices, including precise references to pages, tables, and sections rather than general sources.
• Ability to conduct research on unfamiliar topics from the ground up.
• Native or near-native proficiency in written English.
• Excellent tolerance for detailed structured documentation; constructing the evidence trail forms the majority of the work.
• Previous experience in large language model evaluation, red-teaming, or building benchmarks.
• Background in one or more fields such as reference librarianship, archival research, special collections, investigative journalism, professional fact-checking, OSINT, due diligence, KYC, patent or legal prior-art searches, genealogy, or competitive puzzle construction.

Preferred Skills

• Knowledge of JSON and other structured data formats usage for delivering research findings.

Additional Information

• Work commitment is full-time, 40 hours per week, requiring at least 4 hours overlapping with Pacific Standard Time.
• Compensation is task-based, with payment of $30 per completed research task, each typically requiring around 3 hours on average.