Feature — Scoring & evaluation
Every call scored the same way
When a screening call ends, the transcript is assessed against the job it was for and scored on the KPIs you weight — 0 to 100 each — into one overall figure and a hire recommendation. The strengths, weaknesses, red flags and reasoning come with it, so the number is something you can check rather than something you have to trust.
Free plan, no credit card. Paid plans include a 14-day free trial.
From hang-up to shortlist
- 01
The transcript is read against the role
Scoring does not happen in the abstract. The transcript is assessed alongside the job it was for — title, description, and requirements — and the candidate's own profile, so the same answer counts differently for a role that asked for it and a role that did not.
- 02
Each KPI is scored out of 100
Six KPIs come configured, each carrying a weight that decides how much of the overall score it accounts for. You can replace them per campaign when a role turns on something else.
- Relevant experience — 25%: experience and skills against the job's requirements
- Communication skills — 20%: clarity, articulation, and professionalism
- Enthusiasm — 15%: interest in the role and the company
- Problem solving — 15%: working through a scenario to an answer
- Cultural fit — 15%: alignment with how your team works
- Availability — 10%: whether they can start when you need them
- 03
The weights resolve to one score and one recommendation
The KPI scores combine, by weight, into a single 0–100 figure, and that figure lands in a named band. The bands are fixed, so a 78 means the same thing in March as it did in January and the same thing on one campaign as on another.
- Strong yes — 90 and above
- Yes — 75 to 89
- Maybe — 60 to 74
- No — 40 to 59
- Strong no — below 40
- 04
The reasoning is attached to the score
A number on its own is not reviewable. Each evaluation carries what it was built from, so you can disagree with it on the evidence rather than on instinct.
- Strengths and weaknesses, listed separately
- Red flags — inconsistency, evasiveness, unprofessionalism — or an empty list
- Key findings: experience match, skills demonstrated, communication quality, notable answers
- A reasoning summary explaining how the score was reached
- The per-KPI breakdown behind the overall figure
- 05
The campaign sorts itself
Scored interviews rank against each other, so the top of a campaign is a shortlist you can act on rather than a list you still have to read. Each row carries the candidate, the score, the recommendation, and how long the call ran.
The same rubric, asked of everyone
Every candidate on a campaign is scored against the same KPIs at the same weights, from a transcript of the same questions. That consistency is the point: a first-round score stops depending on which afternoon a resume was read, or which member of the team read it.
Weight what the role actually turns on
The defaults suit a general screen. A shift-based role is decided by availability far more than by problem solving, and a client-facing one by communication; reweighting the KPIs, or replacing them, changes what the score is measuring rather than just what it is called. Individual questions can carry their own rubric where a specific answer needs judging on its own terms.
Scores are evidence, not verdicts
The evaluation is built to be argued with. It cites what the candidate said, separates what was strong from what was weak, and shows its reasoning — so a recruiter overruling a 62 is doing it against the record. Your team's own feedback is stored alongside the AI's, with its own score and recommendation, rather than overwriting it.
A hiring decision stays a human one
Nothing here rejects anyone. Scoring ranks a first round and tells you where to spend your attention; who advances, who is called again, and who is turned down is a decision your team makes and records.
Campaign-level numbers, not just per-candidate ones
Alongside the individual evaluations, a campaign reports how much of it has actually happened — how many candidates are pending, in progress, completed, or exhausted after retries, along with completion and failure rates. It is the difference between knowing a campaign is running and knowing it is working.
- Completion rate across the interviews scheduled
- Failure rate, with automatic pausing once a threshold is crossed
- Candidate counts by state: pending, in progress, completed, exhausted
- Total interviews and call minutes used