Gaden matches every line of the job description against every resume in the stack, in bulk — then hands you the interview: the questions worth asking this person, what each one is really testing, and where the risk is. You walk into the call prepared instead of reading a resume out loud.
Most screening tools count words. A resume that says "maintenance" eleven times outranks the one that ran the department, and the person who did the job under a different job title never surfaces at all. Matching a job description properly means reading it as a list of requirements and asking, of each one, whether this person has actually done it.
In bulk
Drop two hundred resumes at once. Each one is read against every requirement in the job description — not the title, not the keywords: the requirements. Met, partial, or missing, with the line from the resume that decided it.
Partial is the verdict that earns its keep. It is where most good candidates live, and where every keyword tool has to guess yes or no and gets it wrong in both directions.
The interview
Every generic interview question list on the internet is the same list. This one is written for one person against one brief: the partial matches to close, the missing requirement that could end the conversation, and the claim that sounds bigger than the evidence behind it.
Each question carries what it is testing, so you know why you are asking it and what a bad answer means. That is the difference between running an interview and reading a resume out loud.
From the war order
The stack is not the only place candidates come from. Everyone already sitting in the pipeline gets matched against a new brief the moment it lands — the person you sourced in March for a role that fell through is scored against today's job without anybody remembering to look for them.
And because the desk holds the account too — who the client is, what they turned down last time, what they pay — the strategy that comes back is about this client, not about candidates in general.
Not in sourcing. A recruiter who has read the resume twice still walks in without a plan, asks what the candidate did rather than whether they can do this, and finds out in week three of the guarantee period what a better question would have found in minute four.
Worth being plain about: none of this decides who to hire. It reads the brief, reads the resume, and tells you what to ask — a person runs the interview and a person makes the call. Where the law requires human review of hiring decisions, that is not a compliance box here, it is how the thing works.
Keyword screening
Power pre-screen
Give it the job description as written and the whole stack at once. The brief is read as a list of separate requirements rather than a bag of keywords, and every resume is judged against each one — met, partial or missing, with the line from the resume that decided it. Two hundred resumes come back ranked with their reasoning attached, not with a percentage you have to take on faith.
The ones that close the gaps between their resume and this particular brief. That means the claim that sounds bigger than its evidence, the requirement they half meet, and the one they appear to miss entirely — asked in that order, because a deal-breaker belongs at minute four rather than minute fifty. Each question comes with what it is testing, so a weak answer means something.
They are two halves of the same hour. Screening ranks the stack and tells you who is worth your time. Pre-screen picks up there: it tells you what to do with the six who survived. Most tools stop at the ranking, which is the half that was never the hard part.
No, and that is usually where the placement comes from. When a new work order lands, everyone already in the pipeline is rescored against it — the person sourced in March for a role that fell through is matched to today's brief without anyone remembering to go looking. The stack you already own is the cheapest source you have.
No. It reads the brief, reads the resume, and tells you what to ask. A person runs the interview and a person makes the call. Several jurisdictions now require meaningful human review of hiring decisions, and that is not a box being ticked here — the product has no path that rejects a candidate on its own.
That is the case keyword tools fail hardest, and the reason requirements are read one at a time. A maintenance supervisor who was called a "team lead" still supervised a crew, still built the PM schedule, and still shows up as meeting those requirements — with the line that proves it, so you can judge the call rather than trust it.
It comes back with who is worth interviewing, and with the interview.