Decision Room
Bring every attempt and every piece of evidence for a candidate into a single decision view, with the context you need to choose a verdict.
Decision Room
The Decision Room is the single page where you decide on a candidate. It aggregates every attempt the candidate has made in your workspace, the full integrity evidence for each, your screening history, your notes, and the explicit limits of what the assessment can tell you — so a strong result is a strong result, and a weak or assisted one is clearly flagged for your review.
What the page shows
| Section | What it is |
|---|---|
| Identity & cohort | Name, email, roles they have been screened for, and a percentile against the candidate's same-organisation, same-role peers. |
| Verdict history | Every decision event: screening verdicts, recruiter overrides, shortlist changes, and advances. Each event records who acted, when, and (if automated) the policy version. |
| All attempts | Every assessment they have taken, with status, score, and the CETS recommendation. |
| Per-problem evidence | For the current attempt, the integrity and quality signals and the process metrics per problem. |
| Limits of the assessment | A short list of what the assessment cannot speak to, so you know what to ask in the interview. |
| Notes | Your private notes on this candidate. |
| Unavailable panels | Sections this candidate does not have, stated as absences rather than blanks. |
The cohort and percentile
A bare score is undecidable. The same 68 invites every reviewer to apply their
private scale, which is how two people disagree while both believing they applied
the same bar.
The Decision Room shows the candidate's score with a percentile against their
peers — same organisation, same role. The percentile is null when the cohort is
too small to characterise, and the reason (for example, "only 3 peers") is shown
in place of a number. A small cohort is information, not a missing feature.
Verdicts, history, and audit trail
Every change to a candidate's verdict is recorded in the history. Each event records:
- Kind — screening verdict, recommendation override, shortlist add or remove, advance.
- Source —
HUMAN,SYSTEM(mirrored automatically from another part of the product), orPOLICY(decided by the bar, not by a person). - Actor — the user who acted, when the policy decided.
- Reason — the optional reason provided.
- From / to — what changed.
This log is the answer to "why was this candidate rejected in March" a year later.
Limits and what to ask in the interview
Every assessment has limits. The Decision Room states them explicitly rather than hiding them:
- "The candidate was not asked system design."
- "No problem tested concurrency."
- "Video is missing for the first 5 minutes."
A stated absence is more useful than an empty section, because you know what to probe in the next round.
Notes
Add private notes from the page. Notes are visible only to your workspace and only to people who can see the candidate.
What to do next
Video and code replay
Watch the proctoring video, read the video audit events, and replay the candidate's code in sync with the timeline.
Handoff Pack
A focused interview brief that travels with the candidate to the next round: the decision so far, the evidence, the limits, and the questions worth asking.