What is semantic search in recruitment?
Semantic search matches candidates by meaning instead of keywords: "backend lead who can grow a team" finds people whose experience fits that description, even when their profile never uses those words. It replaces boolean strings with plain language and ranks results by fit rather than keyword count.
Keyword systems reward candidates who happen to write like your query. Semantic systems read what the words mean: a "platform engineering manager" can match a search for an "infrastructure lead", because the work is the same even though the words differ.
In practice the gain is twofold: recruiters stop maintaining synonym lists, and strong candidates stop disappearing because their job title was unusual. In Avery, every semantic match also carries the reasoning, so you can check the machine's reading before you act on it.