Remote Web Research Specialist

Turing•Boston, MA

1d•Remote

About The Position

Turing is seeking a Web Research Specialist to join their team in building an evaluation benchmark for frontier AI browsing agents. The role involves designing research problems that even advanced AI cannot solve, despite full web access. This is an investigative research position focused on constructing challenging questions from verifiable facts and providing a detailed evidence trail. The specialist will create natural-language research questions with verifiable answers, along with clues and constraints spanning various fact types. A significant part of the work will be documenting the research process and findings.

Requirements

Demonstrated open-web research ability, including locating primary records and navigating government and institutional databases, archives, registries, and PDF documents.
Precision with sourcing; cite exact pages, tables, and sections, not just homepages.
Comfort researching unfamiliar subjects from scratch.
Native or near-native written English.
High tolerance for structured documentation, as the evidence trail is the majority of the work.

Nice To Haves

Experience with LLM evaluation, red-teaming, or benchmark construction.
Experience in one or more of the following domains: Reference librarianship, archival research, or special collections; Investigative journalism or professional fact-checking; OSINT, due diligence, KYC, or investigative research; Patent, prior-art, or legal-discovery search; Genealogy and records research; Competitive quizzing or puzzle-hunt construction.
Familiarity with JSON and structured data delivery formats.

Responsibilities

Design research problems that state-of-the-art AI cannot solve, even with full web access and multiple attempts.
Start from a verifiable fact and work backwards to construct a question that makes that fact extremely hard to locate.
Provide a complete, auditable evidence trail for each research question.
Produce a natural-language research question with a short, stable, objectively verifiable answer.
Create clues, each independently checkable, spanning multiple fact types (dates, people, places, organizations, works, events, records, quantities) with specific constraints.
Develop a validation record showing the obvious searches run and their results.