Widen the funnel. Judge the person, not the pedigree.
A CV screen rewards who someone already is — the school, the network, the polished résumé. FirstPanel gives every applicant the same structured interview and scores only what they say and do, so more people get a real look and every decision leaves an audit trail.
CV sifts and gut-feel screens quietly narrow who ever gets seen.
Most applicants are cut before anyone speaks to them — filtered on a résumé that broadcasts background, network, and postcode far louder than capability. A name, a school, a career gap, an unfamiliar accent nudges a rushed screener, and the funnel narrows to people who already look like the last hire. It’s rarely deliberate, and it’s almost never documented. By the time you notice the pattern in your pipeline, the decisions that made it are long gone.
Built for the way you hire
The same interview for everyone
Every applicant gets one structured, conversational interview against the same rubric — no fast-tracked referrals, no informal side doors. A consistent process is the foundation of a fairer one.
Blind, content-only scoring
Eight MERIT-8™ agents score what a candidate says and does against job-relevant criteria. Never appearance, accent, name, age, or gender — and never facial expression or emotion, which FirstPanel does not analyse at all.
A policy compiler that blocks bias upstream
Before a single question is asked, the policy compiler checks your criteria and blocks anything unlawful or discriminatory — so bias is designed out of the interview, not caught after the fact.
Continuous adverse-impact monitoring
Selection rates are monitored across groups on an ongoing basis, so you can see how your process performs and act on it — rather than discovering a skew a year and a hundred hires too late.
Built for LL144 and the EU AI Act
Structured scoring, evidence linked to transcripts, and a full audit trail are designed to support NYC Local Law 144 bias-audit and EU AI Act obligations — with the documentation regulators and candidates can ask to see.
A wider funnel, on purpose
Because everyone is interviewed, not just the pedigreed CVs, capable people who’d have been screened out on paper get a genuine shot — and often surprise you.
Reduce bias — and be able to show your work
No tool eliminates bias, and we won’t claim otherwise. What FirstPanel does is make screening structured, content-only, and fully auditable — so it reduces the room for bias and gives you the evidence to prove how each decision was reached.
- Every score traceable to a moment in the transcript
- Adverse-impact monitoring you can review and act on
- Humans make every call — AI never rejects anyone alone
The competencies that predict success
Eight AI agents each own one competency and score it against your rubric — every rating linked to the moment in the interview that earned it. These are the ones that matter most for this kind of role.
Assessed through job-relevant scenarios and reasoning, so demonstrated capability counts more than where it was acquired.
How someone actually thinks through a real problem — a signal that travels across very different backgrounds and CVs.
Clarity and substance of what’s said, judged on content — explicitly not accent, fluency of a first language, or polish.
Sound, consistent judgment probed through situations rather than credentials — the same bar applied to every applicant.
Questions teams ask
Doesn’t AI just bake in the same biases?+
It can, if you let it score the wrong things. FirstPanel scores only job-relevant content — never appearance, accent, name, age, or gender — runs a policy compiler that blocks discriminatory criteria before questions are asked, and monitors adverse impact continuously. We say it reduces bias and stays auditable, not that it eliminates bias.
Does it analyse faces or emotion?+
No. FirstPanel does not perform facial-emotion or expression analysis — it’s prohibited on the platform. Scoring is content-only: what a candidate says and does against the rubric, nothing about how they look.
How does this help with LL144 or the EU AI Act?+
Structured scoring, transcript-linked evidence, and continuous adverse-impact monitoring are designed to support NYC Local Law 144 bias-audit and EU AI Act obligations, and to give you the documentation those regimes expect. It supports your compliance; it doesn’t replace your own legal review.
Won’t interviewing everyone lower the bar?+
It widens the funnel, not the bar. Everyone is interviewed against the same rubric, so capable people who’d have been cut on a CV get seen — while the standard each is measured against stays identical.
Does a human still decide?+
Always. FirstPanel screens and evidences; a named reviewer makes every hiring decision. AI screens, humans decide — always.
See it on one of your own roles
Pick your longest-open requisition. First interviewed shortlist in about two weeks — keep the reports either way.