Importing candidates from External Sources🔗
Candidates selected by the user in the Source UI can be imported into your database. When a user selects one or multiple candidates to import, the CVs/Resumes are automatically parsed and turned into a structured format. We then push the structured format to a webhook. The data structure also includes the user ID who imported the candidate, the name of the external source from which the candidate was fetched, and a reference using which you can fetch the orginal CV/Resume file (if available).
The webhook on your end needs to be built according to the specification of our Automation API. See the candidate_import event. The advantage of building this webhook is that you can also leverage other (future) capabilities of the Automation API without much additional effort.
Deduplication🔗
During the import process, you are responsible for doing your own duplicate check against the candidates already in your database, and handle duplicates. We recommend automating the deduplication process based on unique identifiers such as email and phone. You could also provide a user interface for manually checking duplicates (in case the confidence is lower). But our experience is that users tend to leave those unreviewed and then valuable data does not end up in their candidate database.
Importing with CV/Resume Parsing provisioned by Textkernel🔗
If your CV/Resume Parsing setup was provisioned by Textkernel, we ensure that the structured output for external candidates is identical to the output you would receive when calling the parsing API directly. This includes all of the add-on fields that are enabled for your account.
Importing with self-provisioned CV/Resume Parsing setups🔗
Although Source is not available in our Self-Service Platform, it can work together with the CV/Resume Parser in the Self-Service (Sovren) Platform. When setting up your Source environment, we ensure that Source will use your existing parsing account. Each delivered candidate is a parse transaction in your account and incurs the associated transaction cost.
Because candidate import transactions are triggered from within Source, you cannot indicate per transaction which add-ons you want to enable. Skills normalization is enabled by default because it is required for matching. If you also wish to use other add-ons, such as profession normalization or geocoding, you can use the standalone endpoints provided in the Self-Service Platform to add that information.
Only the V10 parsing output format is supported. If you are not using V10, you will need to upgrade to V10 first.