{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/039aa0f549ea4b01a203a61cf1bb1fb9\" frameborder=\"0\" width=\"1114\" height=\"835\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":835,"width":1114,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":835,"thumbnail_width":1114,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/039aa0f549ea4b01a203a61cf1bb1fb9-980a971fc0efd4a7.gif","duration":300.1711,"title":"Perplexity ReSearcher","description":"Purpose & Impact: \nReSearcher solves a $30 billion inefficiency: 40% of conference-accepted researchers can't afford $4,700 attendance while companies spend $30,000+ per hire searching for them. \nWe built a reasoning graph autonomously matching peer-reviewed papers to job descriptions—85% cost savings while funding conference access. Deployed across major AI conferences, this could fund 5,000+ researchers annually while saving companies $150M in recruitment fees.\n\nTechnical Implementation: Our dual-portal architecture uses Perplexity's sonar-pro model with return_citations: true and search_domain_filter: ['arxiv.org', 'scholar.google.com', 'github.com'] for academic grounding.\nMulti-step reasoning pipeline:\nfetchResearcherProfile() — Real-time Scholar search with authorship validation\nmatchPaperToJob() — Semantic analysis returning 0-100 scores with alignment/gaps\naskAboutResearcher() — Skill extraction with citations\nEvidence mapping — Paper sections → job requirements\n\nBuilt with Next.js 14 and TypeScript, achieving ~5 second matching for 5 researchers via parallel processing (5-10x speedup) with 30-minute caching."}