Caliber

Caliber brings consistency to enterprise interviewing, in two modes.

In companion mode it joins a live interview as a silent participant. The interviewer gets a private side panel that shows how much of the rubric is covered, suggests follow-up questions, scores each answer with the evidence quoted from the conversation and raises bias flags.

In autonomous mode Caliber runs the interview itself through a video avatar, then writes an evaluation that includes a fairness check and a recommended next step.

Around the interview, Caliber scores CVs and builds a shortlist, improves job descriptions, drafts rubrics, searches transcripts and compares candidates side by side. The candidate experience works in several languages.

1. Define. A job description is sharpened and a rubric is drafted from it.
2. Shortlist. CVs are scored against the rubric.
3. Interview. A person interviews with the companion beside them, or the avatar interviews on its own.
4. Evaluate. Every score points to a quote, and the fairness check runs on the result.
5. Compare. Candidates are compared on the same rubric, and any transcript can be searched.

One evaluation engine for both modes

The companion and the autonomous interviewer share one evaluation engine, so a score means the same thing whether a person or the avatar asked the questions.

Scores you can check

Each score carries the passage it rests on, so a hiring team can read the evidence behind every number. Bias flags and the fairness check give the team a second look before a decision is made.

Caliber is for talent acquisition teams and hiring managers at large employers who interview at volume and want consistent, defensible decisions. Interviewers use the companion to stay on the rubric, and recruiting leads use the autonomous mode and the comparison tools to screen at scale.