AI hiring and job matching
Indeed
Indeed operates a job-search and hiring marketplace where people can search openings, post resumes and research employers. Its employer products include Smart Sourcing for finding and contacting candidates, Smart Screening for identifying applicants who meet hiring requirements, and Premium Sponsored Jobs for increasing listings’ visibility with targeting and matching features. Within Smart Sourcing, Sourcing Assistant uses recruiter-written natural-language criteria to identify candidates and invite them to apply, with recruiters controlling outreach volume and approving personalized messages.
Founded in 2004 by Paul Forster and Rony Kahan, Indeed was acquired by Recruit in 2012 and remains a Recruit Holdings subsidiary. Kahan is chairman; Hisayuki “Deko” Idekoba is president and CEO of both Indeed and Recruit Holdings. Indeed’s matching technology learns from qualifications and preferences shared by job seekers and employers. Its sourcing tools look beyond exact keywords to related skills and recent candidate activity.
In June 2026, the company reported 645 million job-seeker profiles and 3.3 million employers across more than 60 countries. Its business includes paid job advertising and sourcing subscriptions. The Indeed app in ChatGPT adds personalized job discovery using a connected profile’s preferences and qualifications, while Indeed’s website and mobile app remain the primary places to apply and connect with employers.
Explore the recordings
Indeed’s supplied recording archive contains one talk. It offers a practical route into MCP App design, rather than a company history. The guidance and implementation examples below describe the recording; they do not establish Indeed’s current product status.
Designing MCP Apps around model exploration
Start with MCP Apps: Give the Model Data, Give the User a UI — Dustin Mihalik. For architecture, follow his recommendation to separate search and data processing from rendering, letting the model explore and filter before displaying selected results. For interaction design, focus on keeping visible UI synchronized with model-readable data and returning user actions to model context. For response behavior, examine his advice to tell the model when results are already rendered, reducing redundant narration. Mihalik uses Claude, ChatGPT, and Indeed’s Career Scout as recorded implementation examples; these are accounts from the talk, not verified descriptions of those products today.
1 talk
Newest first
1 speaker at AIE
Affiliations reflect their AIE appearances, not necessarily current employment.