> ## Documentation Index
> Fetch the complete documentation index at: https://docs.staffer.com/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Sourcing

> Search a large candidate database in your own words, and get results ranked against the job you're actually hiring for.

Search in plain language, not keyword strings, and get candidates ranked for the specific job, not a generic profile score.

## Built on real reach

Staffer searches a large, continuously updated candidate database, not just whoever's applied to you. Every result is scored against that job's actual criteria, so the same person can rank differently for two different roles, because they should.

## Multiple rounds, one conversation

Search rarely nails it in one pass. Ask for a different angle, widen the net, narrow it, and the agent runs another round without losing what you already found. Tell it what to do with the results directly: add the top five, drop anyone outside a timezone, shortlist everyone above a score. No clicking through a list one row at a time.

<CardGroup cols={2}>
  <Card title="Run a search" icon="search" href="/companies/running-a-search">
    How search works, and how to steer it from chat.
  </Card>

  <Card title="Reading match scores" icon="target" href="/companies/match-scores">
    What the number means and where the narrative comes from.
  </Card>
</CardGroup>