For recruiters
Any role. You decide what to measure and what the challenge is. Ready in about 20 minutes.
For candidates
Candidates experience what the job actually feels like. You see how they actually work.
For hiring managers
Every candidate comes back ranked against the bar you set for the role, with proof of how they showed each skill. Compare tradeoffs. Make the call with confidence.
The dimensions every role shares. Each rubric adds the role's own, like the sales rubric on screen above. Every dimension is defined as an observable behavior, so scores come from what the candidate did, not from how they came across.
Whether AI makes them faster — or just louder. We see every prompt, and whether they caught what it got wrong.
The deliverable itself: complete, correct, and at the bar the role actually demands.
How they break down an ambiguous brief, unblock themselves, and keep moving when requirements shift.
Can they explain why — and hold the line on a live call when a manager pushes back?
Who they pull in, what they ask for, and whether feedback changes what they do next.
What they cut when the clock runs out — and whether they flag it instead of quietly missing it.
The honest version. Where the technical platforms already do well — and what only simulating the whole job can show.
Engineering, QA and data roles
Sales, customer success, product, engineering — any JD
Real work from a task library, or your uploaded repo
Real work generated from your JD, no engineering time
Scored on the task: code quality, AI fluency
Encouraged. Every prompt observed — and whether they caught what it got wrong
None — the candidate works alone
AI coworkers who hold the facts they need to ask for
Only with a human interviewer on the calendar
A live call with an AI manager, for every candidate
The task
The whole process: every dimension in the rubric, each score tied to a moment in the recording
Export the full report as a PDF and attach it wherever your process already lives. Direct ATS integrations are in progress — tell us which one you use and we'll prioritise it.
Here is exactly what we touch, and what stays yours.
Candidate data is encrypted on the wire and on disk, and we never sell it.
Every candidate can request deletion from their profile, and we purge the recording, transcript and scores within 30 days.
Results are visible to whoever created the simulation, exportable as a PDF, and shared with nobody else.