How Scoring Works
What Aperture actually measures in an AI interview, and how a candidate's rank is calculated.
Every candidate who completes an Aperture AI interview is evaluated by Aperture's scoring engine, λ-CORE (Lambda Core), a model built to compare candidates against each other for the same role, not against a fixed, one-size-fits-all rubric.
Six behavioral dimensions
Each interview is scored across six dimensions, drawn from the actual conversation rather than anything self-reported on a resume:
| Dimension | What it captures |
|---|---|
| Cognitive Reasoning | Structured thinking, handling complexity, decision-making under uncertainty |
| Domain Knowledge | Demonstrated expertise in conversation, not just claimed on paper |
| Communication | Clarity, precision, the ability to explain complex ideas simply |
| Behavioral Indicators | Patterns of ownership, accountability, evidence of past performance |
| Collaboration | Teamwork instincts, cross-functional ability, listening |
| Adaptability | Comfort with ambiguity, willingness to change course |
Rank, not a raw number
A score in isolation doesn't tell a recruiter much: an 8/10 means different things depending on who else applied. So instead of stopping at a composite score, Aperture produces a confidence range for each dimension and ranks every candidate against everyone else in the same pool for that role.
That ranking is live: as new candidates complete their interviews, the whole pool is re-evaluated, so a candidate's rank always reflects who's actually in the running right now, not a snapshot from whenever they happened to apply.
What shows up on the result
- Rank within the pool (e.g. "3rd of 14"), once there are at least two candidates to compare
- A composite score as a fallback when a pool is too small for rank to be meaningful yet
- A per-dimension breakdown, so a recruiter can see not just where a candidate lands overall, but which of the six dimensions actually drove that result
See the Ashby integration guide for exactly how this result reaches your ATS.
Last Updated September 01, 2026