Exa rolls out semantic search across 1B+ people
Exa just launched a new People Search engine covering over 1 billion public profiles. Powered by fine-tuned Exa embeddings and a hybrid retrieval system, it lets you discover individuals using flexible natural language descriptions instead of rigid keyword matching.
Finding the exact professional you need usually means wrestling with clunky keyword filters on LinkedIn. Exa wants to change that with natural language.
Exa launched its new People Search, allowing users to semantically query over 1 billion profiles across the web.
Semantic discovery across a billion profiles
Instead of relying purely on rigid job titles or company names, the tool runs on a hybrid retrieval system powered by finetuned Exa embeddings.
You can describe the specific background, technical skills, or niche experience you are looking for in plain English. Alongside the release, Exa also published a dedicated benchmark to evaluate people-search performance.
Why it matters
For recruiters, founders, and investors, this could turn hours of manual talent sourcing into a single natural language prompt. It bridges the gap between conversational AI and real-world directory lookup.
Historically, professional networking giants locked profile discovery behind rigid structured forms. Exa is proving that web-scale semantic embeddings can unlock non-traditional, unstructured candidate searches.
We think applying targeted embedding models to people search is a very sharp product move. That said, keeping over 1 billion public profiles up to date without hallucinating career milestones is going to be the real hurdle to watch.
Sources (1)
- @exaailabs(opens in a new tab)Originalx.com