Recrutar AI

Product

A job description goes in. A shortlist with reasons comes out.

Recrutar brings in your applicants and scores every CV against the role’s requirements. AI interviews the strongest, and Autopilot moves candidates to the next round for you. Everything on this page works today.

Your applicants

Start with the applicants you already have

Recrutar brings in a role’s applicants from your ATS, for the months you choose.

Earlier postings of the role, too

Schools advertise the same roles year after year. When you bring in a role, you can include candidates who applied to its earlier postings. They join as applicants to an earlier posting, and you decide whether to contact them.

One candidate, one record

A candidate who applied to five postings is one candidate, not five. You see how many applications came in, how many CVs were read, and any that could not be. Nobody is dropped without being counted.

External search, only when you ask

When your own applicants are not enough, an external search can find professional profiles. It runs only when you ask, for as many candidates as you choose. It keeps work details only: never personal details such as age or gender, a personal email or a phone number. Everyone it finds joins the role as found by external search, and you decide whether to contact them.

Requirements

The role’s requirements, in a list you can edit

When you publish a job description, AI turns it into a list of requirements. A requirement is a must-have only when your advert says it is.

AI checks the job description

As you submit a description, AI suggests clearer requirements: a vague line, a dated one, too many must-haves. You accept or ignore each suggestion.

You edit the list, and the scores follow

Every CV is scored against these requirements, so you decide what they are. Change one, and the CVs are scored again against your version.

Every version is kept

Each change to a job description is saved as a new version, so you can always see the words a score came from.

A colleague can review it without an account

Send a head teacher or hiring manager a link to review the job description. They need no account, and the link expires.

CV scores

A score you can open

Open any candidate to see why they scored as they did, requirement by requirement.

Maths Teacher (Secondary)

312 candidates applied on your ATS · 311 CVs read, 1 could not be read

Example
1Priya RaghavanMaths teacher · applied on your ATSStrong
  • Meets

    Qualified Teacher Status (QTS)Must-have

    “PGCE in Secondary Mathematics with QTS, 2019.”

  • Partly meets

    Experience leading a maths department

    “Coordinated the Year 9 scheme of work across five classes.” The CV suggests this but does not say it, so it can only partly meet.

  • No evidence found

    Experience in an international school

    Nothing in the CV matches this.

2Tomás FerreiraMaths teacher · applied to an earlier posting of this roleGood
  • Meets

    Qualified Teacher Status (QTS)Must-have

    “QTS awarded 2016. Teaching mathematics at KS3 and KS4.”

  • No evidence found

    Experience leading a maths department

    Nothing in the CV matches this.

3Hannah ClarkeCV could not be read, so there is no score. Still on the list.No CV score
18 candidates filtered out, each with the reasonOpen a row to see why

Meets, Partly meets, No evidence found

Each requirement gets one of three answers, with the line from the CV that decided it. A CV that only suggests a skill can partly meet it, never fully.

AI reads, your requirements decide

AI reads each CV. The score is then worked out from which of your requirements the CV meets, the same way for every candidate.

Every CV read once

Each CV is read once and kept, so every later role is scored against the same reading. A CV still being read stays on the list, marked, until it is done.

Filters

Filters use facts about the work

You can only filter on facts about the role, such as the right to work or years of experience. The list is fixed, so nobody can set up a filter on a personal characteristic.

What you can filter on

Right to work, visa sponsorship, location, willingness to relocate, notice period, employment type, education level, institutions, years of experience, licences, a language at a given level, and expected salary. The right to work in a country is a requirement of the role. Being a national of it is not, so it is not a filter.

Missing information never filters anyone out

If a candidate’s details do not say, they stay in. Each filter shows how many candidates it could check, for example: “Recorded for 84 of 300. The other 216 stay in.”

Everyone filtered out is counted

Each candidate a filter removes is listed with the reason, so you can check the filter does what you meant.

Interviews and tests

Questions about the role, scored on the answers

A voice interview, when it suits the candidate

Candidates talk to an AI interviewer from a link, at a time that suits them. It asks about your requirements and follows up when an answer is thin. You read the transcript and the score for each answer.

Every question comes from a requirement

Each scored interview question and each test question is about one of the role’s requirements. A generated question that is not is thrown away. Candidates never see a test’s answers.

A blank test answer counts as wrong

A test is marked on all its questions, so skipping a hard one never helps. Candidates are told this before they start.

Only what candidates say is scored

Never their tone, accent, pace or appearance. The interview uses the microphone only, never the camera.

No interview is lost

Each interview is saved as it happens and marked automatically, even if the candidate leaves early.

Moving candidates

Autopilot does the routine moves

Autopilot moves candidates to the next round for you

Candidates who reach the role’s CV screener minimum score move to the next round. So do candidates who reach the pass mark in an AI interview or a test. Anyone below waits for a person. Autopilot never rejects anyone.

You approve the emails

By default, invitations and rejection emails wait in Ready to send until you approve them. Only an admin can turn this off for a role.

A weak must-have waits for you

If a must-have scores 1 out of 5 in an AI interview, or is not covered, the result waits for a person, whatever the overall score.

Rejecting stays with people

Automatic rejection is off unless you switch it on, and switching it on needs a sample that a person reviews and time to undo each rejection. A round you mark as needing approval never moves anyone on without a person.

Every move is recorded

Each move is recorded with who made it and why, whether a person or Autopilot. Recrutar works beside your ATS.

See it on a role you are hiring for

Bring a job description and a few CVs. Open any candidate and see why they scored as they did.