The AI report
Every candidate, summarized in three minutes.
Verdict at the top, evidence underneath. Open any score and you land on the sentence that produced it.
The AI report
Every candidate, summarized in three minutes of reading.
Verdict at the top, evidence underneath. Open any score and you land on the sentence that produced it.
Per question
- Q4 Sharding a hot table2m 11s9.4
- Q5 On-call escalation1m 38s8.8
- Q6 Idempotent consumers0m 22sflag
- Q7 Cost of the rewrite2m 04s7.6
Q6 answered in 22s after a 32-character paste. Worth a live follow-up.
Verdict
Strong hire, with one question to close
Above the bar on system design and coding. Probe idempotency live before the offer.
Weighted against the rubric you set for this req, not a generic score.
Rubric: REQ-4182 v3
Why it reads this way
The verdict is first because that is what you need first.
Strong hire, with one question to close. Then the reason, then the evidence. You should be able to stop reading at any point and still have made a defensible decision.
Every number is a link, not an assertion.
9.1 on system design means nothing on its own. Open it and you are at the moment they described the sharded write path, with the surrounding thirty seconds of transcript.
Flags are evidence, not accusations.
One paste event of 32 characters at question six is a fact. Whether it matters is your call, and the report is careful never to make that call for you.
What is in it
One record, whatever the candidate did.
Rubric-weighted scoring
Weighted against the rubric you set for this req, not a generic employability score that means nothing across roles.
Per-question breakdown
Time taken and score for each question, so a strong answer given in twenty seconds after a paste is visible rather than averaged away.
Integrity summary
Voice consistency, tab focus, paste events and latency across the whole process, each with a timestamp and the raw evidence.
Side-by-side compare
Two or three shortlisted candidates against the same criteria, so the debrief is about evidence rather than who interviewed most recently.
Client-ready export
A branded PDF your client can forward internally, with the depth of the web report and none of your tooling on show.
Human decision, recorded
The hire or no-hire, who made it and why, stored with the evidence. That is what you want when someone asks six months later.
Questions about reports and scoring
- Where does a score come from?
- Your rubric. Each criterion is weighted by you, and every number links back to the moment in the transcript, the code diff or the panel scorecard that produced it. A score you cannot trace is a score you should not trust.
- Can I compare candidates side by side?
- Yes. Any two or three candidates on a req can be laid out against the same rubric, criterion by criterion, with the supporting evidence one click away.
- Can I send a report to a client?
- Export a branded PDF shortlist. On agency accounts it carries your logo and colours, and Recruit360 is not named anywhere in it.
- How long are reports kept?
- Ninety days on Starter, and for the retention period you configure on Growth and Agency. Deletion on request is honoured across recordings, transcripts and scores together.
Keep reading
Get started
Read the evidence, not forty CVs.
One report per candidate, every score traceable to the moment that earned it.
No card. 1 job free forever. Cancel by closing the tab.