The d.vinci Prescreening is an AI-powered feature in d.vinci applicant management that automatically evaluates incoming applications based on recruiter-defined search criteria. The goal is to provide you as a recruiter with a structured, transparent decision support tool for pre-selecting applications without taking the decision away from you.
For each application, the prescreening provides an overall result (fit) as well as individual evaluations for each search criterion, each with a clear explanation. The final decision on the further handling of an application is always made by a human.
Definitions
Prescreening
Automated preliminary evaluation of applications based on defined criteria before a human reviews the application in detail.
Search Criteria
Requirements for an application (e.g., skills, experience) against which an application is evaluated. Defined by the recruiter, optionally suggested with AI support.
Fit
Result of evaluating an application in relation to the search criteria. Possible values: Strong Fit, Partial Fit, Low Fit, Not Assessable.
Bias Minimization
A technical step preceding the actual evaluation that detects and removes discrimination-relevant information.
Configuration
Prescreening is controlled via two separate permission areas – for creating criteria and for viewing results. This allows clear and separate control over who can define criteria and who can view the results.
Creating Criteria
- Creating and editing search criteria is linked to the existing permission Job Openings | view/edit/create/delete.
- There is no separate prescreening permission – anyone who can edit job openings can also create criteria.
- This automatically limits the group to users who are already authorized to maintain job openings.
- The final determination of the criteria is always made by a human user, even if the AI suggestion feature is used.
Viewing Results
- To see prescreening results (overall evaluation and detailed evaluations per criterion), the system permission Prescreening | View Results is required.
- This permission is included by default in the role Human Resources (Standard).
- It can also be assigned to other roles (e.g., department) if needed.
- This allows precise control over which users may view prescreening results.
Organizational Units and Access
- Access is exclusively role-based.
- Restriction based on organizational units is currently not possible – the feature is active for all organizational levels and units.
Behavior in the System
Creating Search Criteria
- Create a new job opening for which prescreening should be used. You can add prescreening individually to existing job openings.
- Enter a text with the job requirements, e.g., a job description or a profile of requirements.
- This text is sent to Azure AI (GPT-5.1) (see "How is AI used in d.vinci?"). The AI analyzes it and generates suggestions for search criteria.
- Review the suggestions: You can accept, adjust, delete, or manually add any search criteria.
- Alternatively, you can create and edit search criteria completely manually without using the AI suggestion feature.
- You as the recruiter exclusively determine the final list of search criteria. This forms the basis for the subsequent evaluation of incoming applications.
Editing Search Criteria
You can edit, supplement, and refine the list of search criteria at any time to more accurately reflect the match between the requirement profile and the evaluation. You will find this in the job opening under the Prescreening tab.
Triggering Prescreening
- Prescreening is automatically triggered as soon as a new application is received for a job opening that has prescreening search criteria set.
- Processing occurs asynchronously in the background; you do not need to initiate anything manually.
- If the list of search criteria is adjusted later, all affected applications are re-evaluated – except applications with a final status. This reduces unnecessary AI costs.
Results in the Application Overview
- Prescreening results are clearly displayed in the application list (if the corresponding permission is granted).
- Filtering by prescreening fit is possible, with values: Strong Fit, Partial Fit, Low Fit, Not Assessable.
- Sorting by prescreening fit is also possible. Ascending order: Strong Fit → Partial Fit → Low Fit → Not Assessable → Prescreening Pending.
- An aggregated overall result is also calculated.
Detailed View of an Application
- AI-generated evaluations are clearly marked as such in the detailed view (and in the application overview).
- Each search criterion is evaluated individually, including an explanation of why the application fits better or worse.
Data Used
The evaluation especially uses:
- Form data from the application process
- Cover letter
- Resume
- Certificates and diplomas
- Answers to screening questions
- Other attachments submitted by the applicant
Human Oversight and Decision Responsibility
The d.vinci prescreening makes no automated approvals or rejections. It provides an initial assessment as decision support. The final decision is always made by a recruiter or another human decision-maker in the company.
Specifically, as recruiters you can at any time:
- Review, accept, adjust, or delete AI suggestions for search criteria
- Manually add search criteria
- Ignore AI evaluations
- Override results professionally
- View details of the AI assessments
Notes
- The d.vinci prescreening will be a paid feature. More information will follow soon.
- The d.vinci prescreening should not be confused with screening questions.