Describe the person who'd thrive on your team — how they work, the scope they own, the values you hold. Claude matches against rich candidate profiles, not keyword-stuffed PDFs. You see the match reasoning before you see the CV.
Working style, scope, team shape, what success looks like 6 months in.
Candidates surface with score + reasoning, anonymized until you both agree to talk.
Reach out with a tailored note. Their experience claims are pre-verified against their profile.
Their profile and yours. The server can't read either. Even we don't see your hiring data.
Three things every ATS gets wrong, and we built around. Profile-first, reasoning-first, mutual-consent-first.
A title and a bullet list don't describe who'd thrive here. Tell us the working style, the scope, the kind of decisions this person owns. Claude matches that — not your job description against their resume.
Every match comes with a why — attribute-by-attribute. You see "their last three roles shipped customer-facing LLM products to B2B teams" before you see the CV. Lead with reasoning, dig into the resume when the match earns it.
Candidate profiles are pseudonymous in your inbox. Your team's hiring profile is pseudonymous in theirs. When both sides green-light the match, you see each other. No resume mill. No recruiter spray.
You don't post a job and wait. You describe the person, and matches stream in as candidates with that profile shape opt into being considered.
Skip the bullet-list job description. Describe who'd thrive — working style, the team they'd be on, what success looks like in their first six months. The role title is one slot among many.
Candidates whose profiles match yours opt into being considered. They arrive with a fit score and a one-paragraph reasoning. No CV pile, no spray.
Open a match: you see the score, the reasoning, and the attribute-by-attribute breakdown — all before the CV. Claim-traced means every line in their experience summary points back to a verified profile claim.
Their working pattern matches yours on 7 of 8 attributes. Same async default. Last three roles all shipped customer-facing LLM products at companies under 500. The title gap (Manager → Staff IC) is irrelevant — their de-facto scope has been lead-IC for 3 years.
Compose a short intro — the system suggests citation hooks from their profile so your note isn't generic. They get the message under their pseudonym. If they say yes, the masks drop on both sides and you start the real conversation.
You spend hiring time reading reasoning, not sorting through resumes. Every candidate arrives scored, anonymized, and with their experience already claim-checked.
Preview · employer side opens after seeker launch · join the waitlist
Candidates whose profile matches yours opt into being considered. You receive a scored, reasoned match — not a stack of resumes to sort.
Score is computed on working pattern, scope, and verified claims — not on names, photos, schools, or other protected attributes.
Every "shipped X" or "led Y" in their summary points back to a verified profile claim. No more "did this actually happen?" interview question.
A candidate's job search can cost them their current job. A startup's hiring plan can leak strategy. Encrypting one side and not the other only solves half the problem. We do both.
We're onboarding 20 hiring teams in summer 2026. Drop your details and we'll reach out for a 20-minute call to understand what you're hiring for.
Free during private beta. Cohort 2 pricing estimated at €280 / role / month — free while we're onboarding founding teams.
If you've got more, write us — [email protected].
Sourcing tools find people who match your job title and spam them. ATSs filter incoming applications by keyword overlap. We do neither. We match the person you describe against the person each candidate is — working style, scope, values, the things that actually predict thriving on your team. The role title is one slot among many.
Active job-seekers using hunt.work for their own search. They've already built a rich profile of who they are. When a hiring profile matches theirs, both sides get notified. No scraping, no LinkedIn cold-outreach. Candidates opt in to being considered, role by role.
It's encrypted at rest and isolated by per-account row-level policies the database itself enforces. Candidates only see your pseudonym + the working-pattern attributes — your company name surfaces only after they say yes.
Anonymous scoring is on by default and can't be turned off during the match phase. Names, photos, schools, and other protected attributes are not visible to you (or to Claude) until both sides agree to talk. Match scores explicitly avoid those signals.
Today: export matches as structured candidates (Greenhouse, Lever, Ashby) once both sides agree to talk. Bi-directional sync (read JDs from your ATS) is on the roadmap for cohort 2 — please flag this on your access request.
Free during private beta (20 hiring teams, summer 2026). For cohort 2 we're testing a per-role flat fee (€280 / role / month estimated) — no per-candidate / per-message / per-hire spam fees. Founding employers get pricing locked in for 24 months.
20 minutes to write a hiring profile. Matches start surfacing the same day. Free during private beta — founding-employer pricing locked when we open cohort 2.