From 1,000 CVsto 10 talentsin 10 min
Analyses every CV
Every application is read in full against the requirements you set for the role — the first one and the thousandth.
No queue to work through. Applications are analysed as they arrive, not when someone finds the time.
Names removed first. Name, contact details, age and gender are stripped before anything is evaluated.
One standard per role. Criteria are defined once and applied to the whole pool, start to finish.
Volume stays manageable. A role that draws 1,000 applications is no more work than one that draws 50.
Your score
Your score
Your score
Labels who fits and why
Each requirement gets its own label, together with the passage in the application that earned it.
Four plain labels. No, partial, good or perfect match — per requirement, not one opaque score.
Evidence you can check. Every label quotes the lines it came from, so you can disagree with it.
Meaning, not keywords. The same work under a different job title or in another industry still counts.
You set the weight. Mark a requirement must-have or nice-to-have and the reading follows suit.
6+ years professional spellcraft experience, including senior-level ritual ownership
“Ran the Restricted Section rewarding from first survey to final seal, with two junior charmers reporting to me for the length of it.”Strong evidence
Crafts extended reports
For everyone worth a closer look, a written report you read in a minute instead of a CV you read in ten.
A snapshot up front. Career path and experience in a few lines, before you open the document.
Areas to validate. Gaps, unclear claims and contradictions, flagged ahead of the interview.
Questions to ask. Interview questions drawn from exactly what the application leaves open.
Hidden gems. Strong signals a quick skim would walk straight past.
Shortlists top candidates
The pool comes back in order, so the first person you read is the one most likely to be right.
Ranked, never rejected. The order changes; the pool does not. Nobody is turned down automatically.
Strengths at a glance. Skills, experience, credentials and culture shown side by side per candidate.
The decision stays yours. CometHire supplies the reasoning — your team decides who to talk to.
Syncs with your ATS
Results land back where your team already works. Your ATS keeps the pipeline; CometHire does the reading.
Personio and Greenhouse. Connected today, with further systems on the way.
Nothing to migrate. Your stages, your templates and your candidate communication stay put.
Built and hosted in Germany. EU subprocessors only, and a data processing agreement signed up front.
One-click ATS integrations
Syncs natively with Personio Recruiting — no CSV exports, no manual updates.
Connects to Greenhouse via API — evidence labels land straight on the candidate's profile.
Two-way sync with Ashby keeps your pipeline and CometHire evaluations in step.
Works alongside Softgarden's application flow, writing labels straight back into it.
Plugs into Recruitee pipeline stages, so results show up where recruiters already look.
Syncs with BambooHR Hiring — no separate login, no extra screen to check.
TRUSTED BY PARTNERS
Keep your AI workflow
Ask Claude to open a role, show you the strongest applicants or send the next invitations.
Your team lives in ChatGPT? Ask there — same answers, same reasoning behind them.
Le Chat works the same way, for teams that prefer a European assistant.
Frequently Asked Questions
CometHire is screening software for in-house recruiting teams. It reads every application against the requirement profile for a role, gives out an AI evidence label and produces a written rationale for each one. CometHire is built and hosted in Germany and works alongside an existing applicant tracking system (ATS) rather than replacing it.
No. Your ATS keeps the pipeline, communication and scheduling; CometHire handles the evaluation step. It works alongside any ATS. If you don't have an ATS yet, talk to us and we'll tell you honestly whether CometHire is right for you at this stage.
ATS filters and keyword searches match text. CometHire evaluates meaning against a requirement profile: it recognises that a candidate ran a comparable system at a different scale, under a different job title or in a different industry, even when none of the words match. That difference matters most for technical and specialist roles, where the strongest candidates often describe systems and decisions rather than listing tool names.
Three things: consistency, context and accountability. Ad-hoc prompting means criteria shift between candidates and sessions. CometHire applies one defined requirement profile across the whole pool, evaluates in relation to that profile rather than matching text and produces a documented rationale per candidate. It also runs on EU infrastructure with the data processing agreements in place — which a personal chat account does not.
CometHire evaluates applications against the role's requirement profile and shows the evidence behind each rating, rather than scoring against keywords extracted from a job description. It does not automate rejections and it does not conduct interviews — the recruiter decides and CometHire supplies the reasoning. It is built and hosted in Germany, which matters for teams that cannot send candidate data outside the EU.
CometHire is built for GDPR requirements: candidate data is hosted in Germany, only EU-based subprocessors are used, a data processing agreement and technical and organisational measures are available and signed before any real candidate data is processed, and candidate privacy notices are provided. Full detail is in the AI & Data Trust Center.
CometHire is built to sit on top of an existing ATS. If you don't have one yet, talk to us — we'll tell you honestly whether we're the right fit for you right now.
Candidate data is stored in Germany at Hetzner, and all processing takes place within the EU. CometHire does not use customer or candidate data to train AI models, and our AI providers are contractually excluded from doing so. Candidate data is deleted six months after a recruitment process closes. Full detail in our AI & Data Trust Center.
No. CometHire does not hire or reject and it does not produce suitability scores. For each requirement you define, it indicates how much supporting evidence it found — Limited, Partial, Good or High Evidence — and links it to the passage in the application.
Two things. Every application is de-identified before it is evaluated — name, contact details, age and gender are removed, so the model that evaluates does not see personal information; your team still sees the full application. Every candidate is assessed against criteria you defined in advance, not against an unwritten standard that shifts between recruiters or across a long day. Each assessment carries a written rationale, so it can be checked rather than assumed. CometHire does not infer personality or emotion from applications.
We discuss pricing on the demo call, based on how many roles you run and how many applications you screen.
30 minutes. We look at how you screen today, how many roles and applications you handle and whether CometHire would make a difference. If it would, we look at the output on one of your live roles — so you can check the results without any obligation.
Turn your next 300 applications into 10.
See how in 30 minutes.