product · ai pipeline

Eight AI layers.
Every decision, documented.

Not a black box that spits out a number: every stage scores with evidence, explains its reasoning and leaves an auditable trail. You configure the order, thresholds and which stages exist for each job.

how a candidate progresses

Criba CV93
Skills50
Psicométrico78
Entrevista IA71
Copiloto

→ cv_screening: andrea_g.pdf 93/100

→ contradiction_check: 1 alerta

→ copilot.suggest(): pregunta enviada ·

The 8 layers.

01

CV Screening

Reads every CV against the real job requirements: fit, experience, skills and education, with cited evidence and configurable auto-reject threshold.

02

Skills assessment

Technical and situational questions generated for the role — editable — with AI-written response detection.

03

Psychometric tests

56 tests (personality, cognitive, languages, coding…) chosen by AI for the role, evaluated by dimension.

04

Technical test

Take-home or case study designed by AI with weighted evaluation criteria. Manual, AI or hybrid marking.

05

Video presentation

1-3 minutes analysed: communication, language and brand fit.

06

AI interview

Realistic avatar or voice only. STAR questions + real-world job simulations. The candidate acts, not just describes.

07

Human copilot

In the final interview, the AI sits with you: transcribes, suggests, detects contradictions and crosses off your prepared questions.

08

References & report

Automated references with AI analysis and a final report with justified recommendation, risks and onboarding plan.

Your pipeline, at a glance.

Every candidate progresses through stages with their score and footprint. You see where each one is and why.

app.klyria.ai/ofertas/store-manager
Criba CV 38
JB 93
MP 88
PR 84
+35 más
Tests 19
TM 79
SK 76
+17 más
Entrevista IA 9
AL 85
DS 71
Finalistas 4
NW 91
MC 87

real-world job simulations

Don't ask what they'd do.
Put them to work.

The AI recreates the toughest situation in the role — a customer returning without a receipt, a manager pushing for speed, a deal falling through — and the candidate has to act, not describe what they'd theoretically do. You evaluate what actually matters.

  • → retail & sales: objections, complaints, upselling
  • → customer service: escalations and cancellations
  • → management: underperformance, team conflict

simulación · la ia interpreta a la clienta

Clienta enfadada

Quiero devolver este vestido AHORA. No tengo el ticket pero lo compré aquí, ¡lo sé seguro!

Candidata

La entiendo, es frustrante. Déjeme buscar la compra con su tarjeta — ¿la tiene a mano? Y si no aparece, le ofrezco un vale por el importe.

Clienta enfadada

…¿un vale? Bueno. Si me sirve para la colección nueva, vale.

evaluando…

Desescalada
Solución alternativa
Protocolo de devoluciones

anti-fraud

Every answer, watched closely.

Every candidate builds a Trust Score with detailed flags. Penalties are graduated and always visible to the recruiter — the AI flags, you decide.

Tab switchingCopy-pasteAI-generated responsesCross-stage consistencyTrust Score per candidate

the final footprint: 8 dimensions, one candidate

talent pool & nurture

Your talent pool
fills itself.

No good candidate slips away. Anyone who didn't fit a vacancy enters your pool with their consent, and when you post something that does fit, nurture campaigns reactivate them automatically — with open and click tracking.

you don't start from scratch every role · GDPR by design

your talent pool 1.240

"Backend Senior" campaign reactivated

68% opens · 24% clicks · 12 re-applications

Set up your first pipeline in 5 minutes.

Start for free See the interactive demo