Skip to content
AI applicant tracking system

Screen thousands of resumes. Explain every score.

The AI reads every resume against the job. HireRabbit.AI computes every total, so you can show anyone where a score came from.

Resume score breakdown: skill match 84, work experience 76, projects 70, education 62; headline 73, computed by HireRabbit.AI from the four.

AI that does the reading. People who make the call. Every score explained, every decision on record.

Hiring has a volume problem and a trust problem.

Applications are up, recruiting teams are smaller, and few candidates trust the AI tools meant to help.

244

applications per job in 2025, up 111% from 116 in 2022

Source: Greenhouse Hire Standard benchmarks (2026)

+412%

more applications per recruiter, 2022 to 2025

Source: Greenhouse Hire Standard benchmarks (2026)

8%

of job seekers call AI in hiring fair, while 70% of hiring managers trust it

Source: Greenhouse AI Trust Crisis survey (2025)
Why we built HireRabbit.AI

How it works

From a one-line brief to a decision you can explain.

  1. Write the job

    Describe the role in a line or upload a draft. Rabbit writes the description and pulls out the required skills for you to confirm.

  2. Screen every resume

    Upload up to 50 at a time. Each one is parsed and scored on four criteria against the job, not against the pile.

  3. Interview with AI

    Send a link. Candidates consent, then take a real-time spoken interview. You get the recording, transcript and report.

  4. Decide with reasons

    Move people through your own stages. Single moves need a written reason, and key moves email the candidate.

By design

Specific rules, not vague promises.

The product is built around a few hard numbers. Here they are.

4

criteria on every resume

Skill match, work experience, projects and education, each tied to evidence.

50

resumes per upload

Parsed and scored in the background while you keep working.

36

permissions for custom roles

Every checkbox in the role editor gates a real capability.

20

top score with zero required skills

A hard cap in code, whatever the model returned.

AI that does the reading and shows its work.

Three rules run through every part of the product.

  • The AI judges. The platform computes.

    The model judges each resume on four criteria. HireRabbit.AI computes every total, average and count you see. The headline score always agrees with the four scores beside it.

  • The AI suggests. Your rules decide.

    Rabbit, HireRabbit.AI's AI assistant, suggests short reasons; you write the real one. Single-candidate moves need a recorded reason, and auto-reject only runs on a threshold you set and a batch you confirm.

  • Guardrails in code, not prompts.

    A resume with no overlap with the job's required skills is capped at 20, whatever the model says. Candidate names and emails are kept out of the prompt that suggests move reasons.

Irrelevant resumes can't sneak through.

Upload a batch and HireRabbit.AI parses and scores each resume against the job. One with none of the required skills is capped at 20, in code, not in a prompt.
  • Up to 50 resumes per upload, parsed in the background
  • You set the threshold. Resumes below it are filed to Rejected when you confirm the batch. No email is sent, and you can move any of them back.
How scoring works

Senior Python Engineer

Applicant: React developer

PythonDjangoPostgreSQL

HireRabbit.AI shows

20

No skill overlap: capped at 20

A React developer applying to a Senior Python Engineer role: the model returned 82, but with zero required-skill overlap the score is capped at 20.

AI interviews with a report you can defend.

Send a link and the AI Interviewer runs a real-time spoken interview. You get the recording, the transcript and a report whose headline is computed from the per-question scores.
  • Candidates consent before the session starts
  • Integrity flags are for human review and never change the score
Meet the AI Interviewer
AI interview report: four questions scored 82, 78, 64 and 71 with equal weights; headline 74, computed by HireRabbit.AI.

Run the whole process in one place.

A pipeline you design once

Reusable stages between Shortlisted and Selected. Every job gets its own copy, and stage counts always add up to the total.

Pipeline stage counts that add up to the total of 242 candidates, each candidate counted once.

Compare against the job

Skill coverage is measured against the role's required skills, not the applicant pool.

Side-by-side comparison of three candidates against the job's five required skills; GraphQL is thin because only one candidate covers it.

Interviewers see their stage

Assign interviewers to stages. Custom roles build on a catalog of 36 permissions.

An interviewer assigned to the Tech round stage sees only those candidates; every other stage is locked for them.

For the whole team

Built for every seat on the hiring team.

Recruiters

Build jobs, screen batches, run the pipeline and keep every decision on record.

  • AI-drafted job descriptions
  • Batch resume scoring
  • Recruiter-set auto-reject

Hiring managers

See why a candidate ranks where they do before the first conversation.

  • Four-criterion breakdowns
  • Side-by-side comparison
  • Interview reports

Interviewers

Work the stage you own, with notes and reasons, and nothing you do not need.

  • Stage-scoped access
  • Private or shared notes
  • One email when assigned

Admins

Decide who sees candidate data, down to a single pipeline stage.

  • Roles from 36 permissions
  • Instant revocation
  • Invite-only teammates

Black-box scoring vs. a score you can explain.

Some tools let the model assert a number. Others hide the number entirely. HireRabbit.AI shows it and how it was computed.

Black-box scoring compared with HireRabbit.AI
FeatureBlack-box scoringHireRabbit.AI
Where the number comes fromThe model outputs it.HireRabbit.AI computes it from four visible criterion scores.
Irrelevant resumesDepends on the prompt.Zero required-skill overlap caps the score at 20, in code.
RejectionsCan happen with no one looking.You set the threshold. Resumes below it are filed to Rejected when you confirm the batch. No email is sent, and you can move any of them back.
Stage-move reasonsOften none recorded.Required on every single-candidate move. Rabbit only suggests.
See how HireRabbit.AI compares

Frequently asked questions

Does the AI decide who gets rejected?

No. You set the threshold. Resumes below it are filed to Rejected when you confirm the batch. No email is sent, and you can move any of them back. Every other rejection is made by a recruiter, and moving a single candidate requires a written reason.

Where do the scores come from?

The AI judges each resume on four criteria: skill match, work experience, projects and education. HireRabbit.AI computes the headline score from those four, and shows the breakdown right beside it.

What does the AI Interviewer do?

It runs a real-time spoken interview from a link you send. You get the recording, the transcript and a report whose headline score is computed from the per-question scores.

Does movement tracking affect the interview score?

No. Eye, body and hand movement are used only to flag moments for a recruiter to review. Integrity flags never change a score or reject a candidate, and candidates consent before the interview starts.

Who can see candidate resumes?

Only people whose role allows it, plus interviewers assigned to the pipeline stage a candidate is in. Assignment only ever adds access; it never removes access someone already had.

How do I get access?

HireRabbit.AI is in invite-only early access. Request access with your work email. Our team reviews every request and emails you the decision.

See every score explained.

HireRabbit.AI is in invite-only early access. Request access with your work email and our team will review it.