{"schemaVersion":"jobsearcher.job.v1","id":"485033cb06ff39bcfc4d2e58","url":"https://jobsearcher.com/jobs/485033cb06ff39bcfc4d2e58","canonicalUrl":"https://jobsearcher.com/jobs/485033cb06ff39bcfc4d2e58","title":"Creative Intelligence Research Intern (Remote)","description":"SaliencyLab · Remote · Part-time internship · 2 months · Certificate of completionAbout SaliencyLabSaliencyLab is a pre-spend creative intelligence platform. Advertisers upload an ad and get a scored, benchmarked verdict in under 90 seconds, before spending a euro of media budget.Our scores are model predictions validated against real-world public outcomes across 1,200+ ads, and our benchmark database is the moat that makes them credible. That benchmark is what you'll help grow.The RoleYou'll join the founder directly on two missions:1. Grow the benchmark (core of the role)Source real ads from public ad libraries and transparency platforms, following our documented collection processDownload creatives (image and video) and file them following our naming and folder conventionsLog each ad into the benchmark tracker (structured Google Sheet) with its context and publicly available performance data, captured accurately and consistentlyRun collected ads through the scoring pipeline and flag failures (broken uploads, missing previews, pipeline errors)Hit a weekly collection quota, with an emphasis on data accuracy over raw volume, a wrong number is worse than a missing ad.2. Review ads for publication readinessWe are preparing our benchmark dataset for scientific publication. You'll help make it publication-grade:Verify metadata accuracy: does every row match what the public source actually shows?Perform human attribute review of ads against a defined checklist — your ratings feed our inter-rater reliability analysis, a formal part of the researchFlag ads that shouldn't be published: sensitive content, restricted categories, or unclear provenanceClean and normalize tracker data: deduplicate, fix inconsistent labels, document edge casesYou're a Fit If YouAre meticulous. This role lives and dies on data accuracy: you double-check numbers by reflexAre comfortable in spreadsheets (Google Sheets: filters, sorting, basic formulas; you don't need to be a wizard)Are genuinely curious about advertising, you can watch 30 ads in a row and stay engaged enough to notice what makes them differentCan follow a documented process exactly, and speak up when the process seems wrong rather than silently improvisingHave solid written English (tracker notes and review comments are in English)Nice to have (not required):Marketing, communications, psychology, or data-adjacent studiesFamiliarity with digital advertising from coursework or personal projectsWhat You'll GetA certificate of completion detailing your contribution, signed by the founder, plus a reference letter for strong performersDirect exposure to how an AI creative-scoring product is built and validated, you'll see the full loop from raw ad to model prediction to benchmarkYour review work contributes to an academic dataset publication: real research output you can point toDirect mentorship from the founder; no layers, fast feedbackPractical DetailsLocation: RemoteDuration: 2 monthsHours: 15h/week, flexibleCompensation: Unpaid internship, certificate of completion and reference provided","company":"Saliencylab","rawCompany":"saliencylab","isRemote":true,"isActive":false,"createdAt":"2026-08-07T08:15:20.218Z","occupations":[{"code":"13-1161.00","title":"Market Research Analysts and Marketing Specialists","slug":"market-research-analysts-and-marketing-specialists"},{"code":"19-4061.00","title":"Social Science Research Assistants","slug":"social-science-research-assistants"},{"code":"13-1161.01","title":"Search Marketing Strategists","slug":"search-marketing-strategists"}],"industries":[{"code":"541890","title":"Other Services Related to Advertising","slug":"other-services-related-to-advertising"},{"code":"541810","title":"Advertising Agencies","slug":"advertising-agencies"},{"code":"541910","title":"Marketing Research and Public Opinion Polling","slug":"marketing-research-and-public-opinion-polling"}],"jobPosting":{"@context":"https://schema.org","@type":"JobPosting","title":"Creative Intelligence Research Intern (Remote)","description":"SaliencyLab · Remote · Part-time internship · 2 months · Certificate of completionAbout SaliencyLabSaliencyLab is a pre-spend creative intelligence platform. Advertisers upload an ad and get a scored, benchmarked verdict in under 90 seconds, before spending a euro of media budget.Our scores are model predictions validated against real-world public outcomes across 1,200+ ads, and our benchmark database is the moat that makes them credible. That benchmark is what you'll help grow.The RoleYou'll join the founder directly on two missions:1. Grow the benchmark (core of the role)Source real ads from public ad libraries and transparency platforms, following our documented collection processDownload creatives (image and video) and file them following our naming and folder conventionsLog each ad into the benchmark tracker (structured Google Sheet) with its context and publicly available performance data, captured accurately and consistentlyRun collected ads through the scoring pipeline and flag failures (broken uploads, missing previews, pipeline errors)Hit a weekly collection quota, with an emphasis on data accuracy over raw volume, a wrong number is worse than a missing ad.2. Review ads for publication readinessWe are preparing our benchmark dataset for scientific publication. You'll help make it publication-grade:Verify metadata accuracy: does every row match what the public source actually shows?Perform human attribute review of ads against a defined checklist — your ratings feed our inter-rater reliability analysis, a formal part of the researchFlag ads that shouldn't be published: sensitive content, restricted categories, or unclear provenanceClean and normalize tracker data: deduplicate, fix inconsistent labels, document edge casesYou're a Fit If YouAre meticulous. This role lives and dies on data accuracy: you double-check numbers by reflexAre comfortable in spreadsheets (Google Sheets: filters, sorting, basic formulas; you don't need to be a wizard)Are genuinely curious about advertising, you can watch 30 ads in a row and stay engaged enough to notice what makes them differentCan follow a documented process exactly, and speak up when the process seems wrong rather than silently improvisingHave solid written English (tracker notes and review comments are in English)Nice to have (not required):Marketing, communications, psychology, or data-adjacent studiesFamiliarity with digital advertising from coursework or personal projectsWhat You'll GetA certificate of completion detailing your contribution, signed by the founder, plus a reference letter for strong performersDirect exposure to how an AI creative-scoring product is built and validated, you'll see the full loop from raw ad to model prediction to benchmarkYour review work contributes to an academic dataset publication: real research output you can point toDirect mentorship from the founder; no layers, fast feedbackPractical DetailsLocation: RemoteDuration: 2 monthsHours: 15h/week, flexibleCompensation: Unpaid internship, certificate of completion and reference provided","datePosted":"2026-08-07T08:15:20.218Z","dateModified":"2026-08-07T08:15:20.218Z","hiringOrganization":{"@type":"Organization","name":"Saliencylab","sameAs":"https://jobsearcher.com"},"jobLocationType":"TELECOMMUTE","applicantLocationRequirements":{"@type":"Country","name":"US"},"identifier":{"@type":"PropertyValue","name":"JobSearcher","value":"485033cb06ff39bcfc4d2e58"},"url":"https://jobsearcher.com/jobs/485033cb06ff39bcfc4d2e58"}}