AI-authored · Multidimensional scoring

Know who’s actually good at the work.

mindsparx turns your guidelines into a real competency test — AI drafts the items and gold answers, you validate them, and candidates take it through a private, code-only link. Every submission is scored against the rubric that matters for the job.

Sample candidate report

AccuracyAdherenceConsistencyCompletenessEfficiencyCalibration

How it works

01

Paste your guidelines

Drop in the instructions your reviewers are supposed to follow — no template to fill out first.

02

AI drafts the test

An authoring agent generates candidate items and proposed gold answers straight from your guidelines.

03

You validate & publish

Review every item, adjust the rubric if you want, and publish when it's ready.

04

Candidates take it, scored automatically

Each reviewer gets a private link tied to a code — never their name or email — and results are scored the moment they submit.

Everything a real assessment needs

Codes, never names

Reviewers are identified only by an anonymous code end to end — across every report, every export, every screen.

AI-authored, human-approved

Every test item and its gold answer starts as an AI draft — but nothing ships until a human reviews and approves it.

Rubrics you define

Score reviewers on the dimensions that actually matter for your project, not a generic one-size-fits-all scale.

Per-candidate reports

A radar chart and a full item-by-item breakdown for every submission, ready to share or export.

Run it again anytime

Re-test the same content against a new batch of candidates whenever you need to, without rebuilding anything.

Built for real decisions

Automatic flags for incomplete work, low engagement, and near-zero time spent — so a bad submission can't hide behind a good score.

Ready to see who’s actually qualified?