Purpose
Exterview keeps evaluation behavior traceable through controlled configuration, versioned setup, and reviewed changes. The same inputs produce the same result, and you can trace every score back to the rules that produced it.How consistency is maintained
Versioned configuration
Every rubric change creates a new version. It never silently overwrites the last.
Config freeze
A job’s scoring configuration is locked so every candidate in that job is
scored the same way. Prior results stay tied to the version they were scored
under.
Full traceability
Every score traces back to the configuration in effect at the time, and to
the model and prompt version that produced it.
Reviewed change control
Only authorized roles can change configuration, through Governance.
This is the compounding advantage: because your configuration is versioned and
frozen, evaluations stay reproducible as your calibration improves. Results
don’t drift underneath you.
FAQs
Will a candidate's score change if we update our rubric?
Will a candidate's score change if we update our rubric?
Not retroactively. Changes create a new version; prior results stay tied to the version they were scored under.
Can we trace why a candidate scored the way they did?
Can we trace why a candidate scored the way they did?
Yes. Scores trace to the configuration version and rubric in effect at the time.
Who can change the configuration?
Who can change the configuration?
Only authorized roles, through reviewed change control. See Governance.
Related
- Analytics. Roll individual scores up into pipeline-wide trends.
- Governance Setup. Who’s authorized to change configuration, and how it’s reviewed.