MAIA
AI platform for clinical simulations
Your cases. Your criteria. Assessed by AI.
Clinical training cannot
rely solely on standardised patients
Standardised patient programmes are expensive, hard to scale and depend on actor availability. Assessment is subjective and delayed. Faculty spend weeks preparing cases, while students get limited practice and minimal feedback.
Free consultations. Not dialogue trees.
MAIA replaces expensive, hard-to-scale standardised patient programmes with AI that adapts to any group size. Create clinical cases in minutes. Assign them to students. Get personalised assessment against your own criteria, automatically. Faculty retain full control. Students get unlimited practice. Every session generates multi-dimensional feedback.
Three steps. Full cycle.
1. Case creation
Describe the clinical scenario in one sentence: AI generates a full patient profile, medical history, expected diagnostic algorithm and a 10–12 point assessment checklist. Or use the manual 4-step editor for full control.
Every specialty. Every difficulty. Fully customisable. Create 20+ publication-ready cases in one afternoon.
2. Practice with the AI patient
Students conduct realistic, free-form consultations with AI patients. No branching dialogues. No pre-set responses. The patient reacts naturally, reveals information gradually and shifts emotional state depending on the student's approach.
Poor communication leads to frustration. An empathetic approach leads to cooperation. Students feel the consequences of their approach in real time.
3. Automatic assessment
AI assesses every session across four competency dimensions defined by the instructor. Students receive instant feedback: category scores, a checklist with cited evidence from the transcript, analysis of open and closed questions, efficiency metrics and concrete areas for improvement.
Faculty review full annotated transcripts and add personalised comments.
Everything you need. In one place.
AI case generator
Describe a scenario in natural language. MAIA generates a full case: patient profile, system prompt, diagnostic algorithm, categorised checklist and difficulty level.
6 patient temperaments
Calm and cooperative. Anxious and talkative. Impatient and brief. Sceptical and argumentative. Confused and elderly. Emotional and distressed. Each one builds different skills.
Assessment across 4 dimensions
Communication, clinical reasoning, algorithm adherence and quality of recommendations, with intelligent calibration that prevents inflated scores on superficial sessions.
Instructor criteria
No generic rubrics. You set the expected clinical algorithm, assessment checklist and recommended action. AI assesses strictly against your standards.
Question analysis
AI classifies every student question. Breakdown of open and closed questions with concrete examples from the transcript, plus an overall assessment of questioning technique.
Cohort analytics dashboard
Real-time data: competency breakdown, score distribution, trends, most-missed steps and student rankings. AI generates insights and recommends interventions.
Patients that react.
Not scripts that run.
Adaptive patient behaviour
AI patients maintain consistent personality, reveal information only when the right questions are asked and shift mood depending on student behaviour. They do not follow a script, they react.
If the student ignores the patient's concerns, they shut down. If they show empathy, the patient shares more. Every session is different, just like real clinical practice.
Intelligent score calibration
Scores are calibrated dynamically against session quality. Short, superficial consultations cannot produce inflated scores. Hard ceilings ensure clinical rigour: a missing recommendation caps the result at 40%. Missed critical steps cap algorithm adherence at 25%.
The hint system provides context-aware guidance during a consultation with a configurable point penalty, encouraging independent clinical reasoning while providing a safety net.
Built for the leaders in medical education
Deans and programme directors
Scale clinical skills training without growing the standardised-patient budget. Roll out across all departments. Measure outcomes with data, not anecdotes.
Faculty
Define your own assessment criteria. Review annotated transcripts with inline comments. See where your group is strong and where it needs additional support.
Students
Unlimited low-stakes practice with instant, actionable feedback. Repeat cases. Track your progress. Learn from detailed breakdowns of every consultation.
Residency programmes
Standardised assessment for large resident cohorts. Consistent criteria. Documented performance history for every trainee.
Reliable infrastructure for sensitive data
GDPR compliance
Infrastructure hosted entirely within the European Union. Full compliance with personal-data protection requirements.
Multilingual support
Interface and clinical cases in multiple languages. Suitable for international education programmes.
Full audit trail
Every session, transcript, assessment and faculty comment is archived. Transparency and traceability at every level.
Data isolation
Students see only their own data. Role-based access for faculty, administrators and learners.
Piloted at Medical University, Pleven