When you end a station, a second AI examiner reads your whole transcript and marks you against the case's model key points, on the three official domains — each on the 4–8 scale (4 poor fail · 5 fail · 6 pass · 7 good pass · 8 exceptional):
Pass = average ≥ 6. You also get blunt, specific feedback and the key points you missed. Your presence metrics (eye contact, pace, filler words) fold into the professionalism mark.
It's an AI approximation of the real marking scheme — treat the number as directional practice feedback, not a guaranteed exam score. The feedback matters more than the exact figure.
Pick a station and you're in the room. The examiner presents a case and questions you — it reacts to what you actually say, probing gaps one question at a time, and redirects (rather than lectures) if you drift. Speak your answers (tap the mic) or type. 10 minutes per station, or end when you're done.
Presence coach: switch it on and your camera runs a light on-device analysis while you speak — tracking eye contact and head steadiness, the body-language cues real examiners pick up on. Afterwards you get presence metrics (eye-contact %, posture, words/min, filler words) that fold into your professionalism mark.
Watch yourself back: optionally record the station — your camera, your voice and the examiner's — into a video you can replay on the results screen to see exactly how you came across.
🔒 Entirely private — the camera feed and any recording stay in your browser; nothing is ever uploaded or seen by anyone, and the recording is discarded when you leave the page. You turn both on with the toggles below the stations.