Sleak
Platform · Recruiting

Test the work, not the talk about it

Sleak's recruiting add-on lets candidates try on the conversations a role actually requires, then scores each one against the same Standard of Excellence your own team is measured on. Available now as an add-on to your Sleak plan.

Same scenario, everyone

Discovery call · price objection

Your Scorecard
  1. Candidate A84
    Uncovering needs
    9/10
    Handling the objection
    8/10
    Agreeing a next step
    7/10
    “Before we talk about price — what is the current setup costing you in deals you don't close?”
  2. Candidate B71
  3. Candidate C66
  4. Candidate D58
People decide. Every score can be reviewed, overridden or reversed.
How it works

It runs on the engine your team already trains on

Recruiting is a realistic voice simulation with a virtual counterpart, evaluated against a Scorecard — the same one the Sleak Coach uses. If your team already trains on Sleak, the add-on reuses what you have built.

  1. You define what good looks like

    Recruiting uses the same Scorecard and Persona your team already trains against, so the assessment reflects the real job at your company rather than a generic aptitude test.

  2. Candidates try on the role

    A public recruiting page lets each candidate work through a realistic scenario from the role, set in the context of your company. They experience the actual job before they take it — a fairer and more informative first step than a screening call.

  3. Everyone is scored the same way

    Each candidate is evaluated against the same Standard of Excellence, with evidence cited from what they actually said. Results are objective and comparable. High performers can auto-qualify for the next round, so high-volume hiring gets consistent first-round screening without adding reviewer hours.

  4. Turning it on is quick

    Because recruiting reuses the Scorecards and Personas already in your account, you point it at an existing role, publish a recruiting page, and candidates can start. We help you configure the scenarios, calibrate the scoring and set up the human review step.

Fairness and compliance

The same standard for everyone is what makes it fair

Human panels drift: the first impression colours the rest, two interviewers rate the same person differently, and the candidate who looks the part gets the benefit of the doubt. The same scenario, scored the same way for everyone, with evidence cited from what each person actually said, gives every candidate the same shot to show the work — whatever their background or CV.

AI involvement is disclosed

Candidates are told they are interacting with an AI system, and how their performance is assessed, before they begin.

Humans decide

Recruiting informs and ranks. It does not make the hiring decision. The system is built so a competent person can review, override or reverse any AI-generated score, and every decision to advance or reject a candidate stays with your team.

Employers get the tooling to stay compliant

An assessment used in hiring is most likely a high-risk AI system under the EU AI Act (Annex III, employment context). As the deployer, your organisation carries obligations including informing candidates and conducting a Fundamental Rights Impact Assessment. Sleak provides the system documentation, human-review mechanism, candidate notification templates and FRIA supporting materials to make that practical.

Data is handled under EU rules

Sleak acts as data processor under an AVV per Art. 28 GDPR. Primary data residency is the EU (Azure Frankfurt and AWS/Supabase EU). Voice is used for transcription only: no biometric identification and no emotion recognition. Voice audio is sent to a US text-to-speech provider for synthesis and is not stored there. No customer or candidate data is used to train AI models.

Sleak's own ISO 27001 certification is in preparation and not yet obtained. The underlying Azure infrastructure is ISO 27001 certified.

Independent research

Checked by people with no stake in the answer.

r = .63
agreement with expert recruiters, from the transcript alone
TUM School of Management, 2026
4.6 / 5
candidate-rated interpersonal fairness, every justice dimension above 4
TUM School of Management, 2026
Transcript only
no voice, face or emotion analysis in the assessment
by design, per the EU AI Act

Interviews measure interviews

Think about the last great hire you made for a talking role, and the last one that did not work out. There is a good chance both of them interviewed well. That is the hard part. An interview mostly measures how someone handles an interview, and the job is a different conversation entirely: a real customer, a real objection, a real moment where the deal can quietly slip away.

A CV tells you where someone has been. References tell you where they worked. Neither really tells you how they sound when a prospect pushes back on price, a frustrated customer needs calming, or a supplier digs in. So hiring for these roles usually leans on proxies — the polished CV, the confident interview, the strong reference — and trusts that they will carry over once the person is on a live call.

Why a live role-play does not fix it

You could test the conversation directly, but a role-play inside an interview is hard to run well. Not every interviewer can play a convincing counterpart, it takes time you rarely have, and no two candidates get quite the same conversation, so the results are difficult to compare fairly.

The comparison is the part that quietly breaks. Two candidates who faced different objections, from different interviewers, on different days have not taken the same test — and once that is true, the scores that come out of it cannot be ranked against each other, however carefully each one was given.

When the work is a conversation, watch the work

The encouraging part is that the conversation is exactly the thing you can now let candidates show you, before anyone commits.

An airline would never hire a pilot without a simulator check. When the work is a conversation, you can watch the work itself.

Every candidate handles the same realistic scenario, set in the context of your company, and is scored against the same Standard of Excellence your own team is measured on — with the evidence cited from what they actually said. Hiring then rests on demonstrated skill rather than on interview polish, and a capable person who does not interview smoothly gets a fairer shot at the role.

FAQ

Questions people ask before they turn it on.

Everything a hiring team asks before switching it on — and the answers we would give on a call.

The add-on

It is a live add-on to your existing Sleak plan, not a separate product. Because it reuses the Scorecards and Personas your team already trains against, turning it on is quick. Talk to us to add it to your plan.

Recruiting is an add-on to the Sleak plan and reuses the same Scorecards and engine your team trains against, so if you already use Sleak, adding it is quick. New to Sleak? Talk to us and we will set both up together.

Fairness

No. Recruiting produces an objective, evidence-cited assessment against your Scorecard. People make the hiring decisions. Any AI-generated score can be reviewed, overridden or reversed by your team.

Every candidate faces the same realistic scenario and is scored against the same Standard of Excellence, with evidence from what they actually said. That reduces the advantage of interview polish and gives capable candidates a chance to show the work directly. An independent study at the Technical University of Munich found candidates rated this kind of assessment as fair across every dimension measured, and the AI's ranking matched that of experienced recruiters.

They get to show what they can actually do instead of how well they interview, and they experience the real job before accepting it. Capable people who do not interview smoothly get a fairer shot at the role.

Compliance and fit

Recruiting is designed for employment-context obligations: disclosed AI involvement, human oversight of every decision, and deployer support materials including documentation and FRIA templates. Compliance obligations are shared between Sleak (provider) and your organisation (deployer), and we equip you to meet yours.

Roles where the work is the conversation: sales, customer success, leadership and procurement, plus similar conversation-heavy positions where how someone handles the discussion is the job.

Still have questions?

Can’t find the answer you’re looking for? Please chat to our friendly team.