ML Research Talent Map — 50K Researcher Profiles
Scraped and structured 50,000 ML researcher profiles from 5 years of ICLR, ICML, and CVPR proceedings. Searchable by research area via vector embeddings.
Project Tags
No Structured Way to Search ML Research Talent
Five years of ICLR, ICML, and CVPR proceedings contained tens of thousands of researchers, but there was no structured dataset linking them to their research areas, affiliations, or co-author networks.
Sourcing relied on manual keyword searches across Google Scholar and spreadsheets that went stale quickly. Labs were hiring from personal networks rather than the full landscape of available researchers.
50K Profiles Scraped, Structured, and Searchable
Scraped and structured 50,000 ML researcher profiles from five years of top-tier conference proceedings. Each profile includes publication history, co-author networks, and research area tags.
Added a search layer using vector embeddings so recruiters can query by research area or methodology and get ranked results. The dataset is kept up to date as new proceedings are published.