Three answers. One pipeline.
Managers stay managers.
Picked runs the screen, the assessment, and the first-round interview for every role. Your managers see three finalists per role, with full transcripts and a score breakdown. The first time a manager spends time is on the second-round in-person, where it matters.
↓ AI screen and assessment
One bar across managers, set by you.
One rubric framework runs across all roles. Different traits per role family, same fairness posture, same audit. Managers tune within a frame they do not own. The bar stays where the org sets it, not where the manager happens to be that week.
↓ Decision engine
Per-role spend caps. Forecast before publish.
Set a spend cap per role. Default 50 candidates. Picked stops vetting at the cap, you decide whether to raise it. Predicted total spend is visible before you publish each role. Ten roles, five hundred candidates each, capped at the median, costs less than one recruiter retainer.
↓ Role setup