Skip to content
rekruto

AI recruitment screening in Danish and English

The AI asks the follow-up questions a good recruiter would ask.

rekruto reads the application, asks the candidate 3-5 questions per round and re-assesses with the answers as evidence. Every ranking shows its reasoning — and the decision stays human.

Try rekruto

14 days. No payment card.

Mette Jacobsen

Senior Backend .NET, round 2/3

AI assessment

84

Technical depth 88 high
Consulting communication 74 medium
Domain experience 81 high

Evidence

“Designed an event-driven order flow service, ~40k outbound messages/day”

Technical depth · CV, p. 2 · confidence 92%

“Led 4 customer migrations from Oracle to Postgres as tech lead”

Domain experience · Answer, round 2 · confidence 81%

“Explained the trade-off between idempotency and delivery guarantee unprompted”

Consulting communication · Answer, round 1 · confidence 67%

Every line cites its source. Scores are decision support — the recruiter confirms or discounts.

Example values from the Nordic Byte demo tenant.

The loop: from application to decision

One autonomous conversation per candidate, after they apply. The rounds stop when the evidence suffices — or at the position deadline, which always wins. No candidate is ghosted, and a model outage never degrades the loop's operation.

  1. Application

    CV, motivation and portfolio are read against the position rubric.

  2. First analysis

    Every criterion is scored — gaps are flagged as questions.

    score 0.52
  3. Question round 1

    3-5 follow-up questions by email. The candidate answers.

    score 0.71
  4. Re-assessment

    The answers become evidence and lift the coverage.

    score 0.88
  5. Decision

    The ranking with its reasoning lands with the recruiter.

The loop stops when the evidence suffices — or at the position deadline. Always.

The candidate takes part in their own evidence

No other screening product closes the loop with the candidate: chatbots talk about the job, interview AIs score. rekruto asks clarifying questions and re-assesses with the answers — the candidate helps produce their own evidence, which takes the edge off the automated-decision objection.

No drip-feed, no spam

The candidate receives one batch of questions per round — never a stream of single questions. A round that expires unanswered does not automatically open a new one; the deadline sweep is the only path that closes an application without the candidate's participation.

Grounded ranking — not a black box

Every "why" has one authoritative answer. Three layers that can never contradict each other — which the recruiter can open, discount and explain to the candidate.

Why 84?

Criterion scores

A weighted mean over the position rubric, computed deterministically after the model call. No score without matching evidence — coverage drops instead.

Why 88 on technical depth?

Evidence citations

Each criterion cites its claim verbatim, with source and confidence. A quote that does not appear in the document is rejected by the validator.

Why no. 2?

Rank paragraph

One paragraph explaining the placement from the deterministic factor table — and what could change it.

Built for the rules, not around them

rekruto assesses candidates — that makes it a high-risk system under the EU AI Act (Annex III, point 4a). Instead of treating the duties as a burden, we ship them as product features.

No automatic rejection

The lowest score never triggers a rejection or a message. Rejection letters are drafts the recruiter reviews, edits and sends. The score stays hidden until the candidate file is opened.

The human in the loop

The recruiter can reorder the ranking, discount a score entirely and record the reason. Every decision is logged with its reason in an immutable audit trail.

Candidate rights

Human review, access, edit, erase and export are product features on all three tiers — not an Enterprise add-on.

Read about GDPR, the AI Act and candidate rights

150 candidates cost DKK 52,500-135,000 in staff time

Manual screening takes 30-60 minutes of partner time per CV. rekruto reads the documents, asks the questions and collects the evidence — the recruiter reads fewer, better-reasoned files.

3-5

questions per round

100%

decisions kept human

0

automatic rejections

Customer voices — coming

The Nordic Byte demo tenant runs the pipeline with two open positions and five candidates with full assessment history. Testimonials and case studies land here as the first customers reach production.

See the loop on your own position

Create the position, seed the candidates and follow the question rounds from the first email.