How it works
Crosshire reads your CV the way a senior recruiter would, then reads the JD with the same depth. We score the match dimension by dimension, cite every claim back to your CV, and flag anything we couldn’t verify.
Behind the score
Deep Parse
An advanced LLM extracts skills, experience, seniority, and logistics from your CV and the JD into structured cards — not keywords. This step takes 4–6 minutes for a CV, 1–2 for a JD.
Match
Each dimension is scored against the JD with cited reasoning grounded in your CV text. Strengths and gaps are surfaced honestly, with calibrated severity. This is where the verdict comes from.
Verify
Every quoted strength is fuzzy-matched back to your CV. Quotes that don't survive verification are flagged "Listed only — not verified" and never silently kept.
Two sides of the same engine
For candidates
- 01Paste any job description (LinkedIn, recruiter email, anywhere)
- 02Upload or paste your CV (PDF, Word, plain text)
- 03Enter your email to receive the Summary
- 04Decide with confidence: apply, apply with clarifications, or don't apply
For recruiters
- 01Post your JD with free red-flag analysis ("rate not disclosed", "vague seniority")
- 02Drop in candidate CVs from email, LinkedIn, or your ATS
- 03See ranked candidates with cited evidence and pre-screening questions
- 04Download per-candidate briefing PDFs for hiring managers
Fast after your first run
First-time analysis takes 10–15 minutes — because we run a deep parse plus match plus verification across 50,000+ tokens of context. Every match after that is faster. Your CV is parsed once and reused. New JD against your existing CV? Usually 3–5 minutes.
How we stay honest
Every skill cited to your CV. Never fabricated.
Skills we couldn't verify: flagged "Listed only — not verified."
Scores include worst-case ranges. No false confidence.
AI processing disclosed (Anthropic Claude, Google Gemini).
GDPR-native. EU-region data processing where possible.
Output is informational. Not a hiring decision.