Your data and where it goes
A CV is private and never leaves the server. A job advert is public. Everything follows from that.
There is one line that decides everything about how candidate data is handled: a CV is private, a job advert is public. Those two are treated completely differently, and the split is built into the system rather than promised in a policy.
CVs never leave the server
Reading a CV and turning it into a profile is done by a Qwen model running on our own machine. The document, the name, the contact details and the profile itself are never sent anywhere else: not to be processed, not to be stored, and never to train anything. That is why this step runs on our own hardware even though a hosted model would be faster.
Job adverts are public, and are read outside
Judging a job advert is a different matter: the text is public, anyone can read it, and the work is heavy. Two models, Mistral and Bonsai, do that part. What they receive is the public advert and an anonymised profile with the name and contact details already stripped out, never a CV and never a person.
Your candidates are yours
Everything runs on a dedicated server in Europe. Candidate data is visible to the users of your own agency and to nobody else; job adverts, being public data, are shared. Every connection is encrypted, logins can be protected with two-factor authentication and one-time backup codes, and every scan, every verdict and every login is logged so you can always see what happened and when.
Who inside our own company can see a candidate?
Everyone on your side sees the same candidates: the boundary is your company, not the individual recruiter, because a desk that hides its candidates from the colleague covering it on Friday is not a desk. What the levels control is administration — who may invite a new colleague, change someone’s level or remove an account — and not who is allowed to look. People join by invitation only; there is no way to sign up from outside and appear in your list.
Last updated: