How SimuPatient works

From educator-approved case to measurable learning

One governed loop connects case design, a case-grounded AI patient, evidence-linked formative scoring, and faculty review—across text or voice.

Step 1
1Educator
Approve the learning case
Author or refine a synthetic case, teaching notes, and rubric before publication.
Step 2
2Programme admin
Assign the experience
Control access, create cohorts, and assign approved cases to the right learners.
Step 3
3Resident + AI patientCore experience
Interview the AI patient
Take a history, communicate, and request findings naturally by text or voice.
Step 4
4Formative AI
Score the reasoning
Assess diagnoses, differentials, safety priorities, management, and rubric actions against transcript evidence.
Step 5
5Faculty
Review and improve
Review evidence, adjust scores or feedback, improve rubrics, and release the learning outcome.

↺ Faculty findings improve the next case, rubric, and learning cycle

The same structured teaching source grounds the encounter and its assessment.

The simulation engine

One case-grounded AI patient

A responsive simulated patient that stays anchored to the approved case while the learner explores it naturally.

One encounter, two natural modes

Type, use voice notes, or speak hands-free without leaving the consultation or losing its transcript.

Grounded in the case

Responses use defined facts, negatives, findings, collateral content, and disclosure rules.

Reveals facts naturally

Relevant information appears when appropriately explored, rather than exposing the whole case upfront.

Unavailable means unavailable

When a requested result is not defined, the experience reports that boundary instead of intentionally inventing one.

One encounter, two natural modes

Fast typed conversations with the complete encounter preserved for review.

Voice-note or hands-free conversation using the same case context and transcript.

Generative AI can still make errors. Case grounding, disclosure controls, traceability, and educator review are designed to reduce and detect them.

Programme focus

Institutions can combine goals for a bounded pilot or broader formative programme.

History taking
Structured questioning, relevant information gathering, and recognition of gaps.
Diagnostic reasoning
Core workflow
Differentials, must-not-miss conditions, likely diagnosis, and immediate management.
Clinical communication
Clear, empathetic, adaptive communication with patients and defined collateral informants.
Case-based learning
Approved synthetic scenarios for independent practice, cohorts, or facilitated teaching.
Faculty review and feedback
Transcript evidence, rubric calibration, score review, overrides, and feedback release.
Other formative education goal
A tailored formative objective agreed during institutional discovery and readiness review.