How it works

What actually happens, in the order it happens.

The sequence an assessment follows from the criteria somebody wrote to the signature that closes it, a guide to which shape of assessment to reach for, and plain definitions of the words this product uses. Nothing here is a download and nothing here is behind a form.

From “assign it” to “signed off”

A practice mock, a hiring panel and a thesis defence run the same six steps. What changes between them is the stakes, who leads, and whose name is on the last one.

  1. Decide what you are measuring

    Somebody writes the criteria — a name, a description and a weight each. That list is what the grader scores against and what the report is organised by. The instructions the model itself is given are editable too: 130 prompts across 34 areas, overridable for your institution or for one assessment.

  2. The brief is assembled

    The interviewer receives a brief drawn from the candidate’s CV, application or submitted work. Staff can review the source material and any shortened extracts before the session.

  3. They are told what they are sitting, then they sit it

    Participants review the session format, duration and consent information before starting. They can complete recorded answers within the scheduled window or join a live interview.

  4. It is scored against those criteria

    The report shows criterion scores, answer feedback and available speech metrics. Where the brief includes candidate claims, the report records how the answers support them.

  5. The report

    Per-criterion scores and their anchors, the transcript and the recording, strengths, gaps and an improvement plan. From the report itself the person can re-answer one weak question and see the new score against the original, and the whole thing exports to CSV or Excel.

  6. A person signs it off

    Reviewers submit independent scorecards before viewing colleagues’ scores. Configure the required sign-offs and release date to match your review process.

session 4f2a · systems design · round 2 Live
Problem decomposition86/100

Anchor 4 — broke the load path into ingest, fan-out and storage before choosing any component.

Trade-off reasoning72/100

Anchor 3 — named the consistency cost, did not quantify the replication lag it implied.

Communication under challenge91/100

Anchor 5 — restated the interviewer's objection before answering it, twice.

14 min 06 s Transcript attachedrubric: eng-ic4-v3
Claims indexfrom CV · checked in session
  • Led migration of a 40-service estate to Kubernetes

    Substantiated in the interview

  • Owned the on-call rotation for payments

    Partly substantiated — follow up

  • Published two papers on distributed consensus

    Not covered this session

Illustrative, and drawn in the shapes the product's own score sheet and dossier use — the wording is written to be representative rather than quoted from anyone's session.

Which assessment should you run?

They share the interview engine, the rubric machinery and the report, and nothing else. What separates them is whether it happens live, who is leading it, and who has to be able to defend the outcome afterwards.

Recorded mock interviews

A fixed question set answered on the candidate's own time, scored against the competencies you chose.

Reach for it when

Let candidates answer a shared question set on their own schedules.

On their own timeOpens only in its window

Live AI interviews

A real-time voice interviewer that asks, listens and follows up — briefed on this person's CV and the role.

Reach for it when

The follow-up is the point — you want to see what happens when someone is pushed on a claim they made themselves.

LiveThe AI conducts it

Human panels with an AI copilot

Your interviewers lead. A shared rail suggests the next question and scores the last answer as it lands.

Reach for it when

The judgement has to be your people's, and what you want from the software is that everyone is asked and scored the same way.

LiveYour interviewers lead

Oral examinations

Module

Assess a student’s project or subject knowledge with an examiner brief, scheduled sitting and coordinator-reviewed mark.

Reach for it when

What is being examined is the student's own work, and the mark of record has to stay with a named academic.

LiveWindows and attempt caps

Admissions interviews

The interview step of an admissions round, briefed on that applicant's own submission and bound to the committee step that signs it off.

Reach for it when

The interview is one step inside a decision a committee will have to defend, and nothing may reach the applicant before the release date.

Live or AI-conductedBound to the committee step

Recorded statements

Elevator pitch, video résumé or statement of purpose — recorded to a teleprompter and scored against a rubric written for that format.

Reach for it when

There is no interviewer at all — you want a piece to camera that can be scored, versioned and shared.

On their own timeRevocable share link

Evaluation journeys

Compose the stages a candidate or student actually moves through, each with a named owner, a rubric and a gate.

Reach for it when

Run several assessments in order, with a named owner approving each stage before the next opens.

Ordered stagesA named owner per stage

Every one of them ends in the same record. See them side by side

The words on the rest of this site

A guide to the terms used in your interviews and assessment reports.

Rubric anchor
A rubric describes what each score means. The report shows the level reached and the criteria used to assess the answer.
Claims index
Key claims from a CV or application are carried into the interview and reviewed against the participant’s answers.
Quorum
How many distinct reviewers must sign a step off before it completes. One means a single owner decides; anything higher keeps the step in progress until that many people have submitted — which is how a committee decision is made to actually require a committee.
Examiner brief
In an oral examination, the per-student document the examiner is given: built from that student's own submitted work, previewable before the sitting, and marked wherever a source had to be truncated. Available where your institution enables the oral-exams module.
Evaluation journey
An ordered set of stages a candidate or student moves through, each with a named owner, a rubric and a gate. The outcome of a stage — pass, revise or fail — is what unlocks the next one, and the person being assessed sees it as Upcoming, In progress and Completed.
ATS score
A résumé score out of 100 from four equally weighted checks — structure, parse safety, quantified impact and keyword coverage — each with its own itemised findings. The model supplies the per-check judgements; the weighting and the total are computed in our code, so the number can be rebuilt from its parts.
NACE competencies
The eight career-readiness competencies defined by the US National Association of Colleges and Employers: career and self-development, communication, critical thinking, equity and inclusion, leadership, professionalism, teamwork, technology. Interview rubrics, delivery metrics, roadmap progress, profile completeness and ATS score are mapped onto them, and they are the default rubric an evaluation journey starts from.
Embargoed decision
Scores and notes stay withheld until the configured release date. Oral-exam results can be released after coordinator sign-off, after the exam window or kept withheld.
Feature module
A module is a group of features enabled for an institution. Oral examinations require activation; course and class availability depends on the institution’s setup.

Need help choosing a format?

Most pilots begin with one cohort and one module — mock interviews for a final-year group, or a single admissions round. That is enough to see real scores against your own rubric before anything wider is agreed.