Screening compares a set of résumés against one job description and ranks them by how well each evidences what the role asks for.
Running it
Choose the position, upload the résumés or select applicants already in your pipeline, and run. Each is parsed, matched against the position's extracted requirements, and scored.
Résumés are processed in a batch. A file that cannot be read — a corrupt PDF, a password-protected document, an image with no readable text — is reported as a named skip rather than being silently scored as empty. A skip is not a zero.
What the ranking is
A ranking of documents against a description. It reads what each résumé evidences: the skills it names, the seniority it demonstrates, the relevance of the history.
It is not a prediction of interview performance, and it is not a decision. Its job is to order a pile so you read the most promising first.
What it does not do
It does not read anyone's verified skills, their previous interviews with you, or anything about them beyond the document you gave it. It does not weight university, employer prestige, or the gaps between roles.
Duplicates
One person can legitimately appear twice in a batch with two different files. Rows belonging to the same person are recognised and collapsed for anything that acts per person — selecting a candidate selects the person, not the row, and the count reflects that.
Credits
Screening consumes credits per résumé. The estimate shown before you run uses your organisation's own rate, so the number you are quoted is the number you are charged.