A visitor or recruiter submits a job description against your published profile, and the AI produces a verdict on the match. It reads your public content and your private context — but the model that writes what a recruiter reads is never shown your private notes.
Detail
What it sees: both halves of your profile. Your public content is the visible evidence; your private context — the reflections, honest notes, Values tab — grounds the verdict. This is the one feature private context reaches, and it is why the assessment can be honest rather than flattering.
How the boundary holds: the assessment is produced in two steps. The first step reads your private context and outputs only a fixed set of conclusion labels — no sentences. The second step writes everything the recruiter actually reads, working from your published profile, the job description, and those labels. It cannot repeat what you wrote privately, because it was never shown it. An automated safeguard additionally checks the finished assessment against your private phrasing.
The conclusions are a fixed vocabulary, not free-form judgement. The public page at profiled.careers/fit-vocabulary lists every conclusion the assessment can reach — the overall verdicts, the confidence grades, and the strength and gap labels. Nothing outside that list can be concluded about you.
You can see every run. AI Activity in the sidebar logs each fit assessment run against your profile. On Free, assessments are limited to 10 per rolling 30 days; Pro removes the cap.
It is decision support, not a decision. The Terms prohibit recruiters from rejecting a candidate solely on a fit score.
Related
- What do the fit verdict terms actually mean?
- Is my private context ever shown to recruiters or visitors?
- What are the Free limits on visitor chat and fit assessments?
Last verified: 2026-08-18