Dr Andrew O'Malley
Senior Lecturer in the School of Medicine, Head of the Education Research Division, and Deputy Programme Director of ScotGEM.
Read his Substack →Our trust partners
OpenAI Academy
The premise
Communication skills have always been hard to teach at scale.
A medical student needs hundreds of patient encounters to become competent at history-taking. A simulated-patient programme can offer them ten, maybe twenty. OSCE stations measure performance once a year and tell you very little about the practice that produced it.
SimPatient was built inside a medical school, by clinicians who teach, to close that gap. Not by replacing simulated patients (they remain the gold standard for high-stakes assessment) but by giving every learner an unlimited supply of structured, rubric-graded practice in the weeks and months between them.
A note from founder
Dr Andrew O’Malley, founder of SimPatient.
Andrew shares why SimPatient was built: realistic consultation practice for every student, assigned across cohorts, with feedback grounded in the rubrics educators already trust.
Flagship · The Diversity Engine
Most AI-patient platforms have a diversity problem. We have the fix.
Our published research found that standard generative models significantly under-represent darker skin tones in medical imagery (P < .001). So we built the Diversity Engine: a custom model that injects real demographic distributions into the generation pipeline. The gap dropped to P = .04, near-representative output. It now sits inside every SimPatient deployment.
Peer-reviewed
Ensuring appropriate representation in AI-generated medical imagery
O’Malley AS, Veenhuizen MA, Ahmed A. · JMIR AI 2024;3:e58275
Read on JMIR AIThe Diversity Engine in production
Personas are artificially generated. Not real people.
How SimPatient works
Scenario lifecycle.
1
Generate
Build a simulated scenario in a few guided steps, or generate a complete patient from a single line. Set the demographics, the presenting complaint, and how the patient behaves, and the Diversity Engine fills in the rest with a representative persona.
Our story
Started as a research curiosity, built by many disciplines.
Three years ago, SimPatient began as a question inside medical education: could AI-simulated patients help learners practise safer, more repeatable conversations before they met real patients?
Mr Sayed Murad
AI engineering, product design, language equity, and the practical work of building usable tools.
Connect on LinkedIn →
Dr Sandhya Duggal
Clinical communication, medical sociology, health research, and the context around patient conversations.
Connect on LinkedIn →
Dr Xining Wang
Human-computer interaction, technology-enhanced learning, evaluation, and evidence-led design.
Connect on LinkedIn →Research & evidence
Published, peer-reviewed, sitting on the shoulders of giants.
Our work appears in peer-reviewed venues alongside the frameworks our assessment engine is built on.
Med Teach · 2024
Validating AI-generated patient consultations against simulated patient standards
We compared rubric-graded scores between SimPatient sessions and human-marked simulated-patient encounters across two cohorts.
Read paper
MedEdPublish · 2024
Designing assessment rubrics for AI patient simulators
A framework for adapting validated communication-skills rubrics to AI-graded environments, including LCSAS-style anchors.
Read paper
Pre-print · 2025
Cohort-scale practice and OSCE outcomes: early evidence
Initial pilot data from University of St Andrews comparing SimPatient-using cohorts against a control group.
Read paper
From our community
“SimPatient is an impressive example of technology that helps students practise complex professional skills before they enter high-stakes clinical settings. It offers a practical way to support authentic learning, formative feedback, and applied assessment in medical education.”
Assoc. Prof. Mike Perkins, SFHEA, PhDHead of the Centre for Research & Innovation • British University of Vietnam
“The SimPatient platform has been superbly built, with numerous flexible settings the student can modify to explore a particular issue and test their capabilities. The automatic feedback tool, personally styled on the individual student's performance, is a huge plus. I can see students wanting to access it to strengthen their understanding and improve confidence.”
Emeritus Professor Gerry HumphrisClinical Academic Health Psychologist • University of St Andrews
“SimPatient gave me a consistent, reliable, and realistic platform for practising clinical communication, letting me standardise the learning environment so all learners were assessed against the same clinical cues. Students reported greater engagement and increased confidence with unfamiliar cases. I would certainly recommend it to colleagues seeking a scalable, evidence-informed tool.”
Predrag BjelogrlicSenior Lecturer • School of Medicine
“With SimPatient, we gave students an AI patient they could interact with, asking questions and learning the main condition, past medical history, drug history, and social life as part of a full history. It made the whole skills session far more realistic and meaningful.”
Iris CezayirliAssociate Lecturer • School of Medicine
The path in
Bring SimPatient into your programme.
Primary
Book a demo
A 30-minute call with a clinician on our team. We’ll show the wizard, run a live consultation in your preferred mode, and walk through how rubric grading would map to your existing curriculum.
Book a demoLower-stakes
Pilot programme
Run a 4-week pilot with a single cohort. We’ll help you set up your org, import your existing marking scheme, and share a written report at the end measuring usage, learner sentiment, and rubric performance.
Request a pilotAll pricing is per-organisation and includes unlimited learners. Get tailored pricing on the call.