The AI DB Search feature helps you find the most suitable profiles for each new position among the candidates already in your database.
Every time you create a new position, the platform automatically analyses the profiles already in your database and suggests the most compatible candidates, already assessed and ready to be considered, before you reopen the search or republish the job posting.
Why use it
Less time and lower costs per position. Profiles already available in the database are considered before launching new candidate attraction campaigns.
A qualified shortlist from the start. A selection of relevant candidates, ranked by suitability and ready for validation.
A more valuable database. Every application collected in the past can be useful again, ensuring that valuable information is not lost over time.
More objective decisions. Anonymisation during the indexing process helps reduce bias from the outset, while the suitability score provides a shared reference point through a GDPR-friendly approach.
How it works
The process consists of three stages:
1. Candidate indexing
Once the feature has been enabled for the environment, every incoming CV is automatically anonymised and indexed following a privacy-by-design approach. The data is structured according to four key dimensions: experience, skills, education and job title.
2. Semantic matching
When a new position is created with the feature enabled, the system compares the position title and description with the indexed profiles using vector search. It retrieves up to 100 candidates with the greatest degree of overlap across the four dimensions.
3. AI Ranking for the position
The retrieved candidates are reassessed specifically in relation to the open role. Rather than providing a simple score, the AI produces a structured assessment across the four dimensions—experience, skills, education and job title, also considering the risk that the candidate may be overqualified compared with their current role.
The result is a compatibility percentage accompanied by a detailed analysis, providing a shortlist that is ready for validation.
Please note: the search, the retrieval of up to 100 profiles, and the assessment through AI Ranking are two separate, consecutive steps. First, the most relevant candidates are identified; they are then analysed in detail against the specific position.
AI structures and accelerates the process. People remain responsible for the final decision.
How to activate the feature
To activate the feature, follow this path: at the end of the process for creating a new position, go to Position summary > click the Other features tab on the right-hand side of the page > switch AI DB Search to ON.
Please note: once AI DB Search has been activated, it cannot be deactivated for that position.
If the feature is enabled for the client but was not activated when the position was created, the section will still appear within the published position. It will initially be empty and display the “Activate feature” button.
Clicking the button starts the same search and assessment process described above, allowing the feature to be activated at a later stage.
Where to find the results
Once the analysis has been completed, the identified candidates appear within the position, at the bottom of the left-hand column, under the Active Search section.
Three pieces of information are displayed for each candidate:
Last data update. The date on which the candidate’s data was last modified.
Search date. The date on which the search was performed.
AI Ranking. The candidate’s compatibility score for the position, together with the related analysis.
Please note: the section only displays candidates, not any other Datasets associated with their original positions, as different positions may use different Datasets.
Moving a candidate to a new position
If a candidate identified through AI DB Search is considered suitable for the open position, the user can copy them directly into the position. This is a transfer, not an invitation to apply, so the candidate is not prompted to submit an application.
The candidate’s profile information cannot be edited at this stage. Users can only add content, including notes. If the new position requires additional information that is not already available, it can be requested or completed according to the position’s standard workflow.
Data visibility
AI DB Search operates at candidate level rather than at individual application level. This means that all candidate data the recruiter has made visible to other users remains visible within this feature, following the same visibility rules already configured elsewhere in the platform.
Candidate communications
Candidates do not receive any automatic communication simply because they have been identified or moved through AI DB Search. A communication is only sent if:
- the candidate is moved to a pipeline step with an associated automated communication; or
- the candidate is manually invited to apply for the new position.
There is currently no dedicated automation for the specific case in which a candidate is considered through AI DB Search, for example when they enter the Inbox. Automations remain exclusively linked to pipeline steps.
Tip: to manage communications with candidates coming from AI DB Search, we recommend creating a dedicated first step in the pipeline where these profiles can be identified and a specific automated communication can be configured.