A pile of applicants,ranked into a shortlist.

It is an optional assistant that ranks the people who applied to one of your opportunities, from the strongest fit to the weakest, with a short reason for each. It reads the role you posted and each candidate, scores every applicant on a fixed zero to one hundred scale, recommends your top picks and a waitlist of next-best backups, and can move them through your pipeline in one click. It is decision support only. It never makes a hire and never messages anyone on its own. You make every final call.

It is fair by construction, not by policy. Candidates are sent to the model alias-keyed, as candidate_01, candidate_02, and so on, with no name, no email, and no school name. Resume and free-text answers are stripped of contact details, and identity-shaped form answers are dropped entirely. Protected attributes such as race, gender, age, disability, and veteran status are never included, and school prestige is never a ranking input. Recruiter instructions are screened before anything is sent, so the ranking cannot be steered toward or away from any group.

It scores fit against the exact opportunity you posted. It reads your required and preferred skills, your minimum GPA, your eligible majors and minors, your work authorization requirement, and the audience you are reaching, then weighs the candidate stated skills, experience, education, goals, and their answers to your form. The weighting adapts to the type of posting: a scholarship rewards academic merit and the essays, a software role rewards shipped projects and code, and a creative role rewards a portfolio. Each candidate gets an absolute score so people ranked in different runs stay directly comparable.

You tell it how many top picks you want. After ranking, the highest scorers up to that number become your recommended top picks, and the next-best scorers after them become your waitlist, your backups just in case a top pick falls through. Everything below that is still ranked and labeled also ranked. You can adjust the cutoff right in the results, then apply the recommendation in one click: the top picks move to a pipeline status you choose and the waitlist moves to Waitlisted. The pipeline re-checks every move against its rules per candidate, so nothing illegal is ever applied silently.

Only when you tell it to. The AI never changes a status on its own. When you click apply the recommendation, the top picks advance to the status you picked and the waitlist moves to Waitlisted, and that flows through the same pipeline everything else does, so the board, the list, the counts, and the candidate timeline all update together. Every move is logged in your audit trail, and Silent review still controls whether any applicant is emailed.

A deep check tells the assistant to read each candidate documents, their resume and cover letter, or the exact files you select, instead of scoring on their profile and form answers alone. It is off by default to keep runs lean. Either way, every candidate score is cached, so cancelling and re-running never charges twice for someone already scored, and a re-run only pays for candidates who are new or whose application changed. A resume parsed once is reused across the platform rather than read again.

Yes. The match opens a full-screen studio that shows real, live progress as it runs: which candidate it is checking, and the actual things it is weighing for your posting, the required skills, the GPA floor, the eligible majors, the work authorization, and the documents when deep check is on. The studio is durable. You can close it and keep working, and re-open it later to find the same run still going or finished. None of that live progress carries any personal data.

For organizations the SLOS AI candidate match is on the Premium plan and above, and your full-access trial includes it so you can try a real run before deciding. For schools and departments it is part of the institutional plan. Posting opportunities and receiving applications is always free, and the transparent Strong Candidate match that scores every applicant against your role is available without the AI.