Live opening · Posted 7 days ago
At a glance
The key details from the original listing.
Your early-applicant advantage
Live timing from JobBeeper.
About the role
Description supplied by the original job listing.
Hirely is recruiting highly skilled Web Research Specialists for an AI evaluation project with a hiring partner. You’ll help build challenging research benchmarks designed to test the ability of frontier AI browsing agents to find and verify information across the open web. This is investigative research, not content writing. You’ll work backwards from verifiable facts to construct difficult research questions, identify independently checkable clues, and build complete evidence trails that allow every answer to be audited.
What You'll Do
Design natural-language research questions with short, stable, objectively verifiable answers
Start from verifiable facts and work backwards to create challenging research problems
Research unfamiliar subjects from scratch using the open web
Locate primary records and authoritative sources
Navigate government and institutional databases, archives, registries, and PDF documents
Develop multiple independently verifiable clues using dates, people, places, organizations, works, events, records, and quantities
Apply specific constraints to make research problems challenging for AI browsing agents
Document the complete evidence trail supporting each research problem
Record exact pages, tables, sections, and other relevant source locations
Run and document obvious search approaches and record the results returned
Validate that research questions have objectively verifiable answers
Produce structured research and validation records
Work with LLM evaluation and AI benchmark requirements
Who We're Looking For
Master's degree OR 3+ years of relevant professional experience
Demonstrated open-web research ability
Experience locating primary records and navigating institutional or government databases
Strong ability to work with archives, registries, and PDF documents
Exceptional sourcing and citation precision
Ability to cite exact pages, tables, sections, or records rather than simply linking to homepages
Comfortable researching completely unfamiliar subjects
Native or near-native written English
Strong attention to detail and structured documentation
High tolerance for evidence-heavy research workflows
Experience with LLM evaluation, red-teaming, or benchmark construction
Relevant Backgrounds
Experience in one or more of the following areas is particularly relevant: reference librarianship, archival or special-collections research, investigative journalism, professional fact-checking, OSINT, due diligence, KYC or investigative research, patent or prior-art research, legal-discovery search, genealogy and records research, competitive quizzing, or puzzle-hunt construction.
Nice to Have
Familiarity with JSON
Experience working with structured data formats
Experience creating reproducible research workflows
Compensation & Terms
Estimated earning potential of approximately $30 per approved task based on the current task mix
8-week project
Ideally up to 40 hours/week
Full-time availability preferred
Application Process
Apply to the opportunity
Complete the required assessment
Candidates who successfully pass the assessment proceed to onboarding
Begin the project following successful onboarding
Important Requirement: This is investigative research rather than content writing. Strong candidates should be comfortable proving their findings through a detailed, auditable evidence trail and citing precise source locations.
About Hirely
Hirely connects professionals with remote opportunities from hiring partners across artificial intelligence, software engineering, cloud, data, research, and other specialised fields.
Hirely acts as a recruitment and talent-sourcing partner for the hiring companies it represents. Final hiring decisions, onboarding, compensation, and project assignments are determined by the hiring partner.
Work arrangement
No
More openings worth a look
Recently tracked roles with full details and direct application links.