OutLoud MD

How the AI oral board simulator works

Dr. Morgan, the AI examiner in OutLoud MD, runs oral-board stations with you. You reason out loud through the case, and it releases history, results, and imaging when you ask for them. It never hints, corrects, or names the diagnosis, so the clinical judgment stays yours. When the case ends, it scores each answer against per-case criteria and gives you written feedback by domain. It works by voice or by text, and it is practice feedback, not an official or predictive score.

At a glance

  • An AI examiner, Dr. Morgan, runs the case end to end.
  • You reason out loud; it releases findings only when you ask.
  • It never hints or names the diagnosis.
  • You get a score and per-domain feedback after each case.
  • Practice by voice or text, in English.
  • Covers all four station formats: clinical decision-making, prioritization, communication, and procedures with ultrasound.

How a practice session works

A session follows the same arc as a real station.

  • Pick a case. Choose a station format and topic, or generate a fresh case. Each case is grounded in the emergency-medicine literature.
  • Talk through it. Dr. Morgan presents the patient, and you narrate your approach out loud: your first moves, your differential, your plan.
  • Ask for history, results, and imaging. When you request a datum, a test, or an image, it unlocks the matching finding. Nothing is handed to you unprompted.
  • Get a score and feedback. After the case, you receive a score and written feedback broken down by domain, so you can see exactly where to tighten up.

To see what each station asks, read the four stations.

How scoring works

Each answer is scored against criteria defined for that specific case, across the competency domains, and grounded in the medical literature rather than a single pass or fail verdict. The point is to surface where your reasoning was strong and where it was thin, in enough detail to act on before the next case. This is practice feedback. It is not an official ABEM score and it does not predict your result on the real exam. ABEM publishes the current scoring and exam details at abem.org.

Voice or text

You can speak your answers out loud, the way you will on the exam, and Dr. Morgan reads its side aloud so it feels like a real spoken encounter. If you would rather type, you can do that instead, and switch between the two whenever you want. Speaking is the closest rehearsal for the real thing, so most residents practice by voice. Practice is unlimited and free during beta. If you are new to OutLoud MD, start with what OutLoud MD is.

Frequently asked questions

Does the AI tell me the diagnosis?
No. Dr. Morgan never hints at, corrects, or names the diagnosis. It presents the case and releases findings when you ask, and you commit to your own reasoning, just like the real oral board.
How do I get history, labs, or imaging?
You ask for them. When you narrate an action, such as ordering a test or requesting an image, Dr. Morgan releases the matching result. Nothing unlocks until you call for it.
Is the score an official or predictive score?
No. The score is practice feedback against per-case criteria, meant to show where your reasoning was strong or thin. It is not an official ABEM result and does not predict how you will do on the real exam. ABEM publishes current details at abem.org.
Can I practice by voice, or do I have to type?
Either. You can speak your answers out loud and Dr. Morgan reads its side aloud, or you can type. Voice mirrors the real spoken exam most closely.
What is the feedback based on?
Each answer is scored against criteria defined for that case, across the competency domains, and grounded in the emergency-medicine literature rather than a single pass or fail call.

Key takeaways

  • Dr. Morgan runs the full case; you reason out loud and it releases findings on request.
  • It never hints at or names the diagnosis, so the judgment stays yours.
  • Scoring is per-case criteria across domains, grounded in the literature.
  • The score is practice feedback, not an official or predictive result.

Updated September 2026