Comparing Candidates
Stori's compare view puts up to five candidates side by side, organized not by score but by the evidence — actual quotes and video clips from their interviews, mapped against the topics that matter for the job. The AI organizes the evidence; you decide who's strongest.
Why Compare This Way
Most ATSs let you compare candidates on numeric scores. That's a useful summary, but it hides the texture of what each candidate actually said. Stori's compare view is built around the opposite philosophy — surface the raw video and quote evidence, organized by topic, and let you make the call.
The AI does the organization (extracting topics from your job spec, matching candidate quotes to topics, identifying trait moments). It does not do the ranking. You watch the evidence side by side and decide.
Starting a Comparison
Open a job and go to the shortlist. Click the scale icon in the toolbar — that's the compare button.
- Click New Comparison.
- Select 2 to 5 candidates from the modal. Candidates without a completed Stori interview are grayed out — there's nothing to compare without interview content.
- Click Compare. The first run takes 15-30 seconds while the AI builds the evidence map; subsequent loads are cached.
The same scale icon also drops down a list of recent comparisons so you can reload one without rebuilding.
How Topics Are Generated
Topics are pulled directly from the job — its responsibilities, required skills, and other content from the job description. Stori embeds those requirements alongside each candidate's interview transcript and matches the most relevant moments per topic.
That means a comparison for a Senior Engineer job might surface topics like "system design," "team leadership," "incident response" — while the same candidates compared in the context of a Director job might surface "hiring decisions," "organizational design," and "stakeholder communication." Topics adapt to the role.
What the Comparison Looks Like
The result page is organized as a grid — topics down the left, candidates across the top. Each cell shows:
- A video card with the relevant interview clip, timestamped to the moment they spoke about that topic.
- A short quote excerpt for skim-reading without playing video.
- Optional transcript overlay showing more context.
Below the topic grid, a Trait Evidence section shows per-candidate moments tied to specific traits (thinking, discipline, execution, communication) and communication sub-scores (clarity, conciseness, storytelling, self-awareness).
Each candidate stays locked to the same column across every topic. So if Candidate A is in column 1 for "system design," they're in column 1 for "team leadership" and every other topic too. Visual consistency makes the side-by-side scan much faster.
Saving and Recalling Comparisons
Every comparison is saved automatically. Open the scale icon dropdown to see your recent comparisons listed by candidate names — click one to reload it instantly without rebuilding the evidence map.
Comparisons are scoped to the job they were created in.