Evaluate technical candidates.
Without being technical.

Paste a job ad. Get calibrated interview questions in seconds. AI scores the transcript. Your client gets a defensible report.

Built by an IT recruitment agency in Krakow. Used on real candidates.

The problem

Your recruiters are great at finding people. Not at evaluating technical depth.

Your recruiter sits across from a data engineer and has to judge their technical depth. Based on what?

Your client asks why you recommended this candidate. You have notes on a napkin.

The EU AI Act now applies to hiring decisions. Can you prove yours were structured?

How it works

Three steps. No technical knowledge required.

1

Paste the job ad

AI generates calibrated interview questions matched to the role, domain, and seniority level. No templates. Every interview is unique.

2

Run the interview

Use Assay alongside Zoom, Teams, or Meet. Or in person. AI transcribes and scores the answers in real-time.

3

Deliver the report

Your client gets a structured, defensible assessment. White-labeled to your agency. Every rating explained.

Choose the depth

Quick screening or full assessment. Same structured process.

Screening

2 questions. Quick pass/fail.

For initial candidate filtering. Takes 10 minutes. Get a clear go/no-go before investing in a full interview.

  • + Fast candidate triage
  • + Consistent first filter
  • + Pass/fail recommendation
Full Interview

6-10 questions. Detailed report.

For deep candidate assessment. Takes 45 minutes. Produces a complete, client-ready report with per-question analysis.

  • + Per-question AI scoring
  • + Real-time coaching whispers
  • + White-labeled client report

Questions generated from YOUR job ad

Not a generic template library. Every interview is generated from the specific job description, calibrated to the domain and seniority level. Python backend? System design for seniors? Competency-based for HR roles? Assay adapts.

paste job ad → 15 questions in 18 seconds

AI scores what they said, not who they are

No face, no LinkedIn profile, no photo enters the AI scoring. The evaluation is based purely on the candidate's answers. Bias reduction built into the architecture, not bolted on as a feature.

input: transcript only. output: rating + explanation.

Every rating has an explanation

Not just a score. Each question gets a 1-5 rating with a written explanation, key signals identified, and evidence from the transcript. Your client sees exactly why the candidate scored the way they did. EU AI Act compliant by design.

Q3: 4/5 — "Demonstrated strong understanding of..."

Works with Zoom, Teams, Meet, or in person

Assay runs alongside your video call platform. Not a replacement. No vendor lock-in. Your recruiters keep using the tools they already know. Assay adds the structure and the scoring.

Works with any platform. In-person mode for face-to-face interviews.

Built and used atInuits Krakow
"We built Assay because we were embarrassed by our own interview process. Our recruiters are great at finding people. But when a client asked us to justify a technical recommendation, we had nothing structured to show them."

Jeroen Fossaert, Founder, Inuits Krakow

Join the waitlist

We're opening access to a small group of recruitment agencies. Leave your email and we'll be in touch.

Questions? jeroen@inuits.eu