Skip to content
    Performance Intelligence Engine

    The intelligence behind every question you practise.

    PIE is not a chatbot and not a question shuffler. It is a measurement engine that reads how you answer, builds evidence about how you perform, and is the only system that decides what you practise next.

    PIE Neural Core · one STEMI question SIMULATED
    Three dimensional brain model with the parietal, prefrontal and temporal lobes, the anterior cingulate and the cerebellum shown in colour
    • Parietal
    • Prefrontal
    • Temporal
    • Anterior cingulate
    • Cerebellum
    A simulated cognitive process for teaching, using illustrative regional emphasis. It is not a recording of brain activity or a validated map of sequential neural processing. Play the five stages
    Architecture

    From raw answers to one traceable decision.

    Signals PIE observes
    • Answer chosen and correctness
    • Time on question
    • Answer changes and their direction
    • Confidence rating
    • Mode: drill or mock exam
    • Review and flashcard recall
    PIE models
    • Capability
      knowledge by topic, from first attempts
    • Calibration
      how sure you felt against how often you were right
    • Timing
      your pace against each question's usual pace
    • Decision
      answer switching and its effect
    • Learning
      how your results change with practice
    • Sustained performance
      how settled your performance is
    What you see
    • Next best stepBeta
    • PIE Intelligence viewsLive
    • Evidence behind each findingBeta
    • Readiness DNAComing soon
    Confidence calibration

    Knowing when you know.

    Each answer can carry a confidence rating. PIE compares how sure you felt with how often you were right, by topic, so you learn where to slow down.

    What a calibration finding looks likeBeta

    “When you marked answers as Sure in this topic, fewer were right than you expected. Slow down on these before committing.”

    Example wording, not your data. PIE shows a finding only when it has enough evidence, together with how certain it is.

    Principles

    How PIE earns your trust.

    Sole adaptive authority

    One engine chooses your practice. No parallel recommender and no fixed rotation. AI Lab explains. PIE decides.

    Traceable decisions

    Recommendations link back to the answers that caused them.

    Honest uncertainty

    When evidence is thin, PIE says there is not enough evidence yet instead of guessing.

    Your data stays yours

    Each candidate sees only their own evidence.

    Neural Core

    Five stages of clinical reasoning, made visible.

    The Neural Core is a teaching simulation. It follows one STEMI question through five stages so you can see where reasoning can go wrong.

    PIE Neural Core · one STEMI question

    SIMULATED
    Three dimensional brain model with the parietal, prefrontal and temporal lobes, the anterior cingulate and the cerebellum shown in colour SIMULATED
    Demo AMC style MCQDemo, not from the bank

    A man aged 58 has had central crushing chest pain for 40 minutes. He is sweaty and nauseated. His ECG shows ST elevation in II, III and aVF. The nearest PCI centre is 25 minutes away. The regional pathway estimate from first medical contact to wire crossing is 105 minutes.

    What is the most appropriate reperfusion strategy?

    1. APrimary PCI
    2. BThrombolysis
    3. CRepeat troponin in 3 hours
    4. DCT pulmonary angiogram

    Reading the stem. Nothing is highlighted yet.

    Stage 1 · Clinical stimulus The question stem arrives. In this simulation, visual input passes through occipital areas to the parietal lobe.

    • Parietal
    • Prefrontal
    • Temporal
    • Anterior cingulate
    • Cerebellum

    What PIE reads from this attempt

    Demo values
    Accuracy
    Correct
    A chosen on the first attempt
    Answer stability
    0 changes
    No switch between A and B
    Time management
    41 s
    Against a 57 s demo item median
    Confidence calibration
    Sure, correct
    Confidence matched the outcome

    Illustrative values for this one demo answer. They are not statistics about any candidate, and PIE never predicts an exam result.

    A simulated cognitive process for teaching, using illustrative regional emphasis. It is not a recording of brain activity or a validated map of sequential neural processing. The 3D model is built from a de-identified research scan (OpenNeuro ds003787, CC0). Regions are coloured for teaching.

    Simulated for teaching. It never shows your brain activity, and the demo values are not candidate data.

    Questions about PIE

    Is PIE an AI chatbot?

    No. PIE is a measurement and decision engine. AI Lab (beta) can explain your PIE results in plain language using your own OpenAI key, but it never makes decisions.

    Does Zyntra read my brain activity?

    No. The Neural Core shows a simulated model of reasoning for teaching. PIE only uses how you answer questions.

    Will PIE tell me if I will pass?

    No. PIE shows the evidence it has about your practice and how certain it is. It does not predict exam results.