AI Recruiter

No, and it is built so that it cannot. The AI scores candidates and sorts them into a shortlist, a reserve list and not-progressing — but a person approves the interview questions before a job goes live, and anyone with the right permission can change any outcome. When they do, the AI score is kept beside their decision rather than overwritten, and their reason becomes part of the record. If a hiring decision is ever questioned, you can show exactly what the AI said, what the person decided, and why.

Every part of every score cites the exact text from the CV that produced it, quoted verbatim. If the AI produces a quote that does not actually appear in the CV, that evidence is discarded rather than shown to you — so a fabricated claim never reaches a hiring record. The report shows concerns as well as strengths, because a report that only lists the positives is not honest and is not useful.

The scoring AI is explicitly instructed to judge only evidence of ability against the requirements you wrote, and to ignore name, gender, age, marital status, nationality, religion, photograph, and the prestige of a university or employer. Because every score cites its evidence, you can check that for yourself on any candidate rather than taking it on trust. A candidate's photo may be shown to your human reviewers — so the person meeting them in the final round has a face to the name — but the scoring never sees it, and you can turn that display off entirely for a fully blind review. AI-assisted hiring is regulated differently by country, so for specific compliance questions in your market, talk to our team.

We will not pretend otherwise: a determined candidate can get help, and no screening tool can fully prevent that. This is why the product is positioned as screening rather than as the decision. It narrows hundreds of applications down to a shortlist worth a human conversation — it does not replace that conversation.

Never. Candidates get a status page showing the stage their application is at, and an outcome email in plain language. They are never shown a number, a bucket name, or how they compared with anyone else. Telling someone a score for a decision about their livelihood is neither kind nor defensible.

They are told. Every ending — shortlisted, reserve or not progressing — sends the candidate an email, and outcome mail is deliberately delayed so a rejection never lands seconds after someone applies. Delivery is tracked too: if an email bounces, that is flagged on the candidate, so silence caused by a bad address is never mistaken for a candidate who ignored you.

HRM manages the people you already employ — attendance, payroll, leave, shifts, records. AI Recruiter handles the stage before that: finding and screening people you have not hired yet, and it stops the moment someone is hired. Many companies run both. If you are unsure which you need, tell us whether the problem is hiring new people or managing existing staff.

Image CVs are first-class. We accept PDF, JPG and PNG on the upload form (Word and text formats were dropped deliberately, because we need to be able to lift the candidate's photo from the CV for the interview identity check, and a Word CV re-exports to PDF trivially). A photographed or scanned resume is read directly by a vision model rather than treated as empty — in several of our markets a photographed CV is normal, not an edge case. Vision reading is a super-admin setting because it costs more per candidate than reading text; when it is on, a scan is read and then scored the same as any other CV. When it is off, or on the rare occasion the file is genuinely unreadable, the application is held for a human to look at and shown in its own group on the report rather than buried among the rejections — we never guess an unreadable CV into a rejection.