AI Examiner Feedback
Good systematic approach to cardiac history. Consider asking about family history of IHD earlier. Your ICE exploration was excellent — keep using open questions for patient concerns.
How it works
From completed station to full grading report in under 30 seconds.
Complete a station
Finish your consultation with an AI patient. Your entire conversation is automatically transcribed word-for-word with timestamps.
AI analyses transcript
Your transcript is compared line-by-line against an examiner-calibrated rubric built from real OSCE marking schemes.
Domain scores generated
You receive separate scores for Data Gathering, Clinical Management, and Interpersonal Skills — the same three domains used by GMC examiners.
Detailed feedback
Every rubric item is annotated: what you said that earned marks, what was missed, and exactly how to improve next time.
Why structured grading accelerates improvement
Most medical students practise OSCE stations without any meaningful feedback. You run through a scenario with a friend, they say "yeah, that was pretty good," and you move on. The problem is that self-assessment in clinical skills is notoriously unreliable — research consistently shows that students who perform worst are also worst at judging their own performance. Without structured, objective feedback after every attempt, you're practising blind.
The GMC OSCE uses three distinct scoring domains, and understanding them is critical. Data Gathering measures whether you asked the right questions in the right way — a structured history with appropriate open and closed questions, relevant systems review, and focused examination findings. Clinical Management assesses whether your diagnosis, investigations, and management plan are safe, evidence-based, and appropriate for the clinical context. Interpersonal Skills evaluates empathy, active listening, shared decision-making, and whether the patient felt heard and respected throughout the consultation.
Our grading rubric is calibrated against real OSCE marking schemes. Each scenario has a detailed rubric with weighted items across all three domains, built from examiner mark sheets and refined against thousands of practice sessions. This isn't a generic AI opinion — it's a structured assessment that maps directly to how you'll be scored on exam day. When the rubric says you missed "exploring the patient's ideas, concerns, and expectations," it means that specific item on the real mark sheet would have been scored zero.
There's a well-documented gap between knowing the clinical content and being able to perform it under exam conditions. You might know exactly what questions to ask for chest pain, but under the pressure of an 8-minute timer with a nervous patient, you forget to ask about family history or rush through the social history. Structured grading exposes these performance gaps — the things you know in theory but drop in practice — so you can target them specifically in your next session.
Consistent, objective feedback creates a dramatically faster improvement curve than informal peer assessment. When every session produces a grading report with the same structure, you can track exactly which domains are improving, which rubric items you consistently miss, and whether your overall trajectory is heading towards a pass. This data-driven approach replaces guesswork with evidence, turning hundreds of practice sessions into a measurable progression towards exam readiness.
What your grading report includes
Overall score
A single percentage score for the entire station, weighted across all three domains to match real OSCE scoring.
Domain breakdown
Separate scores for Data Gathering, Clinical Management, and Interpersonal Skills so you know exactly where to focus.
Rubric items checked
Every item on the marking rubric is listed with a clear pass/fail indicator showing what you covered and what you missed.
Transcript highlights
Key moments in your consultation are highlighted — the questions that earned marks, tagged to the specific rubric item they satisfied.
Missed elements
A clear list of rubric items you didn't cover, with guidance on what you should have asked or said and when in the consultation to do it.
Improvement suggestions
Targeted, actionable advice for each domain — not generic tips, but specific changes based on what happened in your session.
Pass/fail indicator
A clear signal of whether your performance would have met the pass threshold on the real exam, calibrated to GMC standards.
Session comparison
Side-by-side comparison with your previous attempts on the same scenario, showing exactly what changed between sessions.
Domain trends
Track how each domain score evolves over time across all your sessions — spot patterns and measure long-term progress.
Frequently asked questions
Our grading system is calibrated against real OSCE marking schemes and refined through thousands of practice sessions. While no automated system is identical to a human examiner, our rubric-based approach produces consistent, structured feedback that closely mirrors how examiners score on the day. The key advantage over human feedback is consistency — the AI applies the same standard every single time, which makes it ideal for tracking improvement across sessions.
Data Gathering (DG) assesses your history-taking and examination skills — did you ask the right questions, in a logical order, using appropriate technique? Clinical Management (CM) covers your diagnosis, differential, investigations, and management plan — is it safe, evidence-based, and appropriate? Interpersonal Skills (IPS) evaluates rapport, empathy, active listening, and patient-centred communication. All three carry equal weight in the real exam, and our grading reflects this.
Focus on your weakest domain first. If your IPS score is consistently low, spend time practising empathic statements, signposting, and checking patient understanding. If DG is your weak area, work on structuring your history more systematically. Review the 'missed elements' section of each report — these are the specific items to target. Many students find it helpful to re-attempt the same scenario after reviewing feedback, aiming to pick up the items they missed.
We provide a pass/fail indicator calibrated to GMC standards for each station. The exact threshold varies by station type and difficulty, just as it does in the real exam. However, we recommend focusing less on the pass/fail line and more on consistent improvement across domains. A student who scores 60% and is trending upward is in a better position than one who scores 70% but is plateauing.
The rubric adapts to the station type. History-taking stations weight Data Gathering more heavily. Counselling stations emphasise Interpersonal Skills and shared decision-making. Acute stations assess time-critical Clinical Management decisions. Ethics stations evaluate your ability to navigate professional dilemmas. Each station type has a tailored rubric that reflects what examiners actually look for in that specific format.
Your Analytics dashboard shows your scores in the context of overall performance benchmarks, so you can see how your domain scores compare to the cohort average. This helps you understand whether a 65% in Data Gathering is a strong score or one that needs work. All comparisons are anonymised and aggregated — you'll never see individual student data, and no one else can see yours.