An interview scorecard is a fixed 1-5 scale with a written definition at each level, so two interviewers score the same answer the same way. Enter the role, the rounds and the time per round to get one with behavioural anchors at each level, so interviewers score the same answer the same way. Free, no signup.
Interview notes record what happened. A scorecard records a judgement against a fixed standard, which is the thing you can compare across candidates and defend later. Notes alone are why debriefs turn into whoever remembers the most vivid moment.
A 1-5 scale without definitions is a mood ring. "3" must mean something specific — for a data engineer, perhaps "describes a pipeline they built and the failure it survived, but not why they chose that design". The generator writes an anchor for every level of every criterion, which is the part teams skip and then wonder why scores are inconsistent.
Scores written an hour after an interview drift toward overall impression. Scoring each criterion as it is covered keeps the evidence attached to the judgement, and makes it obvious when a panel never actually probed a criterion they scored.
Every scorecard system degrades as volume rises. An AI interviewer returns a completed scorecard per candidate with evidence linked to the recording, so the hundredth interview is scored to the same standard as the first.
Related: Rubric generator · Interview question generator · AI resume grader