The screening interview, ranked.
Pull the candidates we surface through the same five-dimension rubric the shortlist ranks them on. The interviewer is a live, in-browser transcriber — your utterances land tagged to the dimension the probe was crafted to probe. The warehouse reads the row on submit; the morning shortlist knows what to do.
- Ownership
- Collaboration
- Adaptability
- Leadership
- Alignment
What’s persisted per candidate
A single JSON row stub.
Full transcript blob (one span per probe), per-dimension regex scores, and the probes used at capture time. Indexed for the future shortlist endpoint — no migration needed.
- 8 probesacross 5 dimensions — ownership, collaboration, adaptability, leadership, alignment.
- Live transcriptionvia the browser’s Web Speech API — or type if the mic isn’t available.
- Audit trail— matched regex cues per dimension land in the row, not a re-scored stranger.
Probe 1 / 8
Tell me about a time you owned a project end-to-end. What did you do when it stopped being fun?
Tap “Start recording” to dictate, or type below.
Per-dimension preview
Recomputed live from your transcript — exactly what we persist at submit.
- Ownership0 / 10
No cues matched yet for ownership.
- Collaboration0 / 10
No cues matched yet for collaboration.
- Adaptability0 / 10
No cues matched yet for adaptability.
- Leadership0 / 10
No cues matched yet for leadership.
- Alignment0 / 10
No cues matched yet for alignment.
Your transcript will appear here.
As you answer each probe, we’ll tag the span you’ve just finished to the dimension it probes. Five dimensions, eight probes, one shortlist-ready row.