AI functions in Teamio

AI candidate relevance assessment – information for companies

An overview of what the feature is for, how it works, how you control it, and what to watch out for. It also serves as a reference if a candidate asks you how their data is processed. Information on AI features for candidates is available on this page.

What it’s for

For every candidate you select, you get a quick overview of how their profile fits the position, including a percentage match. Treat it as a first orientation, not a decision.

Key facts

• The system pulls the criteria from the job ad. You can’t enter a custom criterion manually — it’s changed by editing the job ad.
• It’s triggered manually. You choose which candidates, in which recruitment campaign, you want to evaluate.
• It’s switched on for the whole organization. It’s activated by users with the HR manager role.
• A human decides. The output is a reference for the recruiter, not a verdict on the candidate.

How it works

The assessment runs in three steps.

1. Job ad analysis
The AI reads the text of the job ad and pulls out the assessment criteria itself. It assigns an importance to each one. The criteria always come from the job ad (more on this in How to change what the system evaluates).

2. Candidate analysis
The AI goes through the available materials — the CV, answers from the pre-selection questionnaire, and internal notes on the candidate — and compares them against the position’s criteria.

3. Evaluation
For each criterion, the AI determines whether the candidate:

• meets it,
• partially meets it,
• doesn’t meet it,
• can’t be assessed from the available data (information is missing from the materials).

Each criterion also comes with a short explanation of what the AI based its assessment on. From these results, the system calculates an overall percentage match between the profile and the job ad — the more criteria met, and the more important they are, the higher the percentage. Key criteria have the biggest influence on the result.

What you’ll see

A candidate card contains:

• a percentage match score with the position,
• an estimate of the candidate’s seniority relative to the role,
• a list of criteria each marked and with a short justification,
• a summary of strengths and risks,
• suggested questions worth discussing with the candidate.

How to change what the system evaluates

The system pulls the assessment criteria from the text of the job ad — they aren’t entered manually. This means:

• You can’t add a custom criterion outside the job ad. The system doesn’t evaluate against anything entered outside the position text (e.g. “prefer a certain gender”).
• To change what’s evaluated, edit the job ad. Specify in the text what matters for the position and how much. Then have the system generate a new set of criteria from the current version of the job ad — the assessment will follow it.

Criteria always reflect the job ad.

Who turns the feature on and who can see it

• The feature is switched on and off at the whole-organization level.
• It can be turned on or off by a user with the HR manager role.
• Once switched on, the feature is visible to all users with the HR manager role in your organization. Everyone with access to it is informed that it’s turned on.
• Turning the feature on or off is logged — it’s traceable who did it and when.

When the assessment runs

The assessment doesn’t run automatically. It’s always up to the user who and where to evaluate.

• You select specific candidates in a specific recruitment and trigger the relevance assessment for them.
• The system doesn’t evaluate all candidates on its own.

What to watch out for

The output is a recommendation, not a verdict. You always have the final say.

• Always check the output. For key criteria, verify what the AI based its assessment on.
• Output quality depends on input quality. The more specific the job ad and the more complete the materials, the better the assessment. Questionnaire answers take precedence over CV data.
• It assesses how well experience matches the position, not the candidate’s personality.

If a candidate asks how it works

You can confidently explain the following to a candidate:

• The system compares their CV, questionnaire answers, and any notes against criteria derived from the text of the job ad they applied to.
• For each criterion, it shows whether they meet it, partially meet it, don’t meet it, or whether it can’t be assessed from the materials — and it always states what this was based on.
• The results produce a percentage match between their profile and the position.
• It only assesses how well their experience matches the position — not personality, age, gender, or origin.
• The result is a reference for the recruiter, not a decision. A human always decides on the candidate’s progress.

Details on data processing are available on the Information for candidates page — you can share this directly with the candidate.

Information on the processing of personal data for advertisements (template)

Download sample text for job offers with information on the processing of applicant data.