Healthcare / Explore
Healthcare / Explore
AI-Enhanced Debriefing for Medical Simulation
Timeline, key moments, and questions for the debrief, minutes after the scenario. Audio only or audio and video. AI running on site: your data never leaves your institution.
01 / The Solution
Debriefing is where simulation turns into learning, and it works best when the facilitator pairs observation with genuine curiosity about the learner’s reasoning. The harder the scenario, the harder it is to hold that stance: hours of recording, emotions in the room, a checklist that says what was omitted but not why.
AI-Enhanced Debriefing puts an AI co-facilitator beside the faculty. Running on your own infrastructure, it captures what people say and, if you choose, where they look, how they interact, and signals from wearable augments. It aligns everything with the scenario timeline, checks your protocols and checklists item by item, flags the moments worth exploring, and prepares a neutral, data-informed question for each, in the spirit of debriefing with good judgment, advocacy-inquiry, and PEARLS. It reads the data, not the person: the facilitator decides what to examine and leads the conversation. Developed through field research with Jenny Rudolph and Roxane Gardner of the Center for Medical Simulation, Boston.
Three capture levels: audio only (room microphones, no cameras), audio and video, and wearable augments (heart rate, HRV, arousal) where appropriate and consented. Works with the AV you already run, in any browser. Recordings, transcripts, and analytics stay in your network; the model is tuned to your protocols, your language, and your rubrics; aggregated data on cohorts, stations, criteria, and readiness are yours, for education, patient safety, and clinical governance.
02 / How It Works
Run the scenario
Realistic scenarios with manikins, standardized patients, or conversational avatars; real teams, real behavior. Voice and, if you choose, attention, interaction, and physiological signals are captured as it unfolds, on site.
Let the AI observe
Signals are aligned with the timeline. The system marks the moments that stand out, checks your protocol items, and prepares a question for each: evidence for the facilitator, never a score for the learner.
Debrief with good judgment
The facilitator reviews the moments, picks what matters for the learning goals, and asks, within PEARLS, advocacy-inquiry, or your own framework. Learners explain their reasoning, compare alternatives, and decide what to do differently.
Govern with your data
Criteria missed by cohort, station, and period; readiness across a rotation; training needs by team. Aggregated in your institution’s own data and kept there. A validation pilot measures agreement between the AI’s key moments and your instructors’ observations before you scale.

