Outpatient clinic
Meet your patient
Dana Mitchell, 46
“Bleeding between my periods.”
Chart: vital signs
About 15 minutes · two visits in one sitting · saved without your name, for testing
Clinical Assessment of Reasoning & Empathy with AI
Your AI Patient Simulation
Outpatient clinic
Meet your patient
“Bleeding between my periods.”
About 15 minutes · two visits in one sitting · saved without your name, for testing

Talk with a mystery patient
About 15 minutes · your conversation is not saved

An assigned patient for the study
Not open yet
You are scored on how you communicate as much as on your clinical reasoning, and she responds to how you treat her.
| Clinical reasoning | History, exam and tests, differential, diagnosis and plan |
|---|---|
| Emotional attunement | Noticing and responding to how she feels |
| Listening | Letting her tell her story and showing you understood it |
| Explanation and shared decisions | Clear words, checking understanding, and deciding together |
| Culture, identity and circumstances | Asking about her life and building it into the plan |
Your browser will ask to use your microphone. Headphones help her hear you clearly. Nothing you say is saved after you close this page.
Prepared encounters skip the voice session and go straight to scoring, so you can check the feedback quickly. They use Dana's case.
These apply to the next encounter you start, standard or open practice. A voice cannot change in the middle of a session, so switching means starting again. The speaking settings are directions given to the voice model, not audio controls, so results vary.
Study sessions and Demo encounters are saved with their transcript, event log, feedback and any comments. Open Practice conversations are not saved. The usage history lists every encounter of any kind, with no content. A transcript file holds the conversation only, so it can go to a faculty rater; a full record adds a short feedback summary and the comments.

Optional. Your answers help improve CARE-AI. Please do not include your name.
This feedback is generated by an AI model from the transcript. It is for practice and is not a grade.