GADEN SCREENING ROOM · POWERED BY AEL

One job description in. A ranked, evidence-backed shortlist out.

The rubric comes from your job description itself, not a generic template. Every score cites the evidence behind it, borderline candidates go to human review, and the profiles come out client-ready.

Hire Gaden See how it scores

RUBRIC FROM THE JD · EVIDENCE ON EVERY SCORE · HUMAN REVIEW ON THE BORDERLINE

WHAT COMES BACK

A shortlist with its reasoning attached.

Four tiers, ranked. Not a black-box number — every candidate carries the line from their own resume that earned the score.

GADEN SCREENING · RN, ICU · 28 RESUMES · RUBRIC FROM JD · 2m 11s

STRONG FIT

FIT

CONSIDER

NOT QUALIFIED

TOP OF THE LIST · CANDIDATE 04 · BLIND PROFILE

Level I / II ICU experience4 yrs · “charge nurse, 24-bed MICU”

Required certificationCCRN · current through 2027

Night availability“nights preferred”

Commute to site38 min · flagged, not disqualifying

Gap Jun 2024 – Jan 2025unexplained · ask on the call

EVERY LINE ABOVE CITES THE RESUME IT CAME FROM · NOTHING INFERRED WITHOUT A SOURCE

WHY YOU CAN ACTUALLY TRUST IT

Three answers to the three things recruiters do not trust about AI screening.

Where did the criteria come from? Your own job description.

Gaden does not score against a generic checklist of buzzwords. It reads the JD you actually wrote, derives the rubric from it — the certifications that role really needs, the years that matter, the setting it happens in — and screens against that. Change the JD and the rubric changes with it.

rubric from the JDrole-specific weightingno generic checklist

RUBRIC DERIVED FROM YOUR JD

ICU experience, Level I or IIweighted heavily

CCRN certificationrequired

Night shift availabilityrequired

Charge or precepting historypreferred

Commute under 45 minsoft signal only

Did it make that up? Every score cites its evidence.

A number on its own is worthless — you cannot defend it to a client and you cannot check it yourself. So every score Gaden gives carries the line from the resume that produced it. If it cannot find the evidence, it does not claim the point; it flags the gap and tells you to ask on the call.

evidence cited per scoregaps flagged, not guesseddefensible to a client

SCORE WITH ITS SOURCE

Strong fit — because the resume says so, not because a model felt it.

ICU depth“charge nurse, 24-bed MICU, 4 yrs”

Certification“CCRN, exp. 03/2027”

Leadership“precepted 6 new grads”

Employment gapno evidence — ask, do not assume

What about the ones on the line? They go to a human.

The dangerous failure in screening is not a wrong ranking — it is a good candidate quietly auto-rejected. Gaden routes borderline cases to human review instead of deciding for you, and can produce blind profiles so a client judges the substance before the name. When you overrule a score, that correction calibrates the next run of the same role.

human review on the borderlineblind profilesclient-ready outputlearns your corrections

SENT TO YOU, NOT REJECTED

Candidate 11strong clinically, thin on ICU

Candidate 17cert lapsed 2 months ago

Candidate 22career changer, unusual path

You moved Candidate 22 to Fit last month on a similar role. Gaden remembers, and weighs career changers differently on this one.

THE DIFFERENCE

Keyword filters reject people. Screening evaluates them.

The resume parser that has been in every ATS for twenty years is still just matching strings. That is why good people never make it to your inbox.

KEYWORD MATCHING

Rejects on a missing word.

Matches strings, not meaning
Drops anyone who phrased it differently
Gives a score with no explanation
Auto-rejects the borderline silently
Never learns what you actually want

GADEN SCREENING

Reads it the way you would.

Builds the rubric from your own JD
Understands equivalent experience
Cites the evidence for every score
Sends the borderline to you, not the bin
Calibrates on your corrections

QUESTIONS PEOPLE ACTUALLY ASK

The short answers.

How does AI resume screening work?

Gaden reads your job description and builds the scoring rubric from it — the requirements that role actually has, not a generic checklist. Each resume is then scored against that rubric, and every score carries the evidence it came from, so you can see the line in the resume that produced it.

Can I trust an AI to screen candidates fairly?

Gaden is built so you never have to take its word for it. Every score cites its evidence, borderline candidates are sent to human review rather than auto-rejected, and profiles can be produced blind so a client reviews the substance first. Where it is unsure, it says so instead of guessing.

How do I screen a lot of resumes at once?

Load a batch of up to ten resumes and Gaden parses each one, including PDFs, then scores them all against the same rubric and ranks them. You get one shortlist with the reasoning attached rather than ten separate opinions.

What is a candidate shortlist that is client-ready?

A shortlist you can forward without rewriting it — each candidate summarised against the role, strengths and gaps stated plainly, and the evidence for both. Blind versions omit identifying details so the client assesses the fit before anything else.

Does the screening get better over time?

Yes. When you correct a score or move a candidate up or down, Gaden keeps that correction and applies it to the next run of the same role, so the rubric converges on how you actually judge that job. It is part of the Autonomous Employee Loop — the same brain that sources, screens, contracts and collects.