Comparable evidence
One frozen version, one timer, per-topic breakdowns — comparison is like for like.
Talent
Talent runs hiring campaigns on the same assessment engine: one frozen version, invitations by email, every candidate on the same timed task, and review decisions you can audit later.
Seven steps from open role to recorded decision.
Organisation admins manage recruiters; candidates get a narrow role that only reaches their own assignments and results.
Questions are human-approved first, then published as an immutable version with its own time limit, pass mark and release policy.
A campaign freezes the assessment version it started with, so a later edit cannot change what candidates were measured on.
Invite an existing verified candidate member, or send an email invitation. An emailed link expires at the earlier of 14 days or the availability end, and only its hash is stored.
Server-set deadline, per-answer saving, resume support and a confirmed submission — identical rules for every candidate.
Recruiters see each candidate's latest submitted attempt side by side, can leave a review note, and record hold, shortlisted or rejected. Every decision keeps an actor, a note and a timestamp.
Under a manual release policy, candidates see nothing until staff release that attempt; the interface never leaks a score early, and never shows answer keys to candidates.
One frozen version, one timer, per-topic breakdowns — comparison is like for like.
Decisions and notes are tenant-scoped and appended to history, so a later reviewer can see who decided what and when.
Candidates see their own assignment, attempt and released result only — never another candidate's data.
Sandbox only: plan upgrades in this deployment use sandbox payment gateways. A browser redirect never activates a plan — the server verifies the payment first.