AI candidate sourcing with quoted evidence
Hyre-Inc's AI candidate sourcing turns a plain-English role description into a ranked list of people. It maps your competitors and same-industry companies, searches public web pages, company team pages, GitHub and your past applicants, and only shows people whose fit score rests on quoted evidence. It never scrapes LinkedIn or buys contact data.
Updated:
How sourcing works, step by step
You start by describing the role the way you would explain it to a colleague, for example "senior backend engineer, Go or Rust, payments experience, EU timezone, open to contract". Hyre then works through five steps on its own:
- Map. It builds a competitor map from your role: direct competitors, same-industry companies, the client side and the tool side.
- Search. It searches public sources for people who work, or used to work, at those companies.
- Resolve. Mentions of the same person across different pages are merged into one profile, so you never review the same person twice.
- Evidence. Every fit score rests on lines quoted from public pages, so you can read why someone fits.
- Rank. People are ranked by location tier and fit. Only people at 70% fit or higher show by default.
Where the people come from
Hyre searches the public web, company team pages, GitHub and your own past applicants. Past applicants matter more than most teams expect: someone who was a near miss for one role last year can be the strongest match for the next one, and they already know your company.
You stay in control of where Hyre looks. Each role has location tiers (primary, then tier 1 to 3) and a "never search here" list, and locations on that list are skipped entirely. Recruiters can also add people by hand, with or without an email address, for example a referral from a colleague. Everyone, however they were found, sits in the same round and moves through the same five steps: Source, Screen, Test, Interview and Hire.
Quoted evidence instead of a black-box score
Most sourcing tools give you a number and ask you to trust it. Hyre shows its working. Each fit score is built from short quotes taken from public pages, such as a team page that names someone's role or a project they shipped, and the quotes sit next to the score.
That changes how a recruiter spends their time. Instead of opening ten tabs to check whether a profile is real, you read two or three quoted lines and decide. If the evidence is thin, the person does not reach the 70% threshold and does not show as a result. You can change that threshold per role when you want a wider or narrower list. When you disagree with a score, you can still move the person forward by hand, and that move is logged with your name and reason.
Sourcing at a glance
| Topic | How Hyre-Inc handles it |
|---|---|
| Input | A role description in plain English |
| Sources | Public web search, company team pages, GitHub, your past applicants, people you add |
| Fit threshold | 70% by default, configurable per role |
| Location control | Primary location, tiers 1 to 3, and a never-search list |
| Duplicates | Mentions of the same person are merged into one profile |
| LinkedIn scraping | No |
| Bought contact data | No |
| Outreach | Drafted by Hyre, approved by a recruiter, sent from your own mailbox, at most 40 a day |
What Hyre-Inc will not do
Hyre does not scrape LinkedIn, does not buy contact data and does not harvest email addresses. We think hiring tools that rely on those shortcuts put both the candidate and the employer in a bad position, and they make it harder to explain to anyone where a profile came from.
Outreach follows the same idea. Hyre drafts a personalised email sequence for each person, and a recruiter reviews and approves it before anything is sent. Emails go from your own connected mailbox, capped at 40 a day, and a sequence stops as soon as the person replies, opts out or moves stage. Opting out takes one click, and that person is never contacted again. Replies sync into the Hyre inbox, so the conversation stays next to the profile and its evidence.
Questions about sourcing
Do you scrape LinkedIn?
No. Hyre does not scrape LinkedIn, does not buy contact data and does not harvest email addresses.
Where do the people come from?
From your role description, Hyre builds a map of your competitors and same-industry companies, then searches the public web, company team pages, GitHub and your own past applicants. You set location tiers and a "never search here" list. Every fit score rests on quoted evidence from public pages, and only people at 70%+ fit (configurable) show as results. Recruiters can also add people by hand, with or without an email.
How does outreach work?
Hyre drafts personalised email sequences, and a recruiter reviews and approves them before anything is sent. Emails go from your own connected mailbox, capped at 40 a day. A sequence stops when the person replies, opts out or moves stage. Opting out takes one click, and that person is never contacted again. Replies sync into the Hyre inbox.
Can we change the 70% fit threshold?
Yes. 70% is the default for showing someone as a result, and you can raise or lower it per role. People below the threshold are not shown by default, and anyone you move forward by hand is logged with your name and reason.