Initialize parameters and constants
Initialize data buffers for multiple timeframes
Initialize historical performance record

For each new bar:
    For each timeframe in {Fast, Medium, Slow}:
        Update price buffer
        Compute AbstractMetric_A[timeframe]  // e.g., some transformation of price history

    Compute AbstractFeatureSet1 from AbstractMetric_A and other transformations

    RawSignal ? AbstractDecisionFunction1(AbstractFeatureSet1)

    ModelInput1 ? AbstractFeatureSet1
    [ProbLong, ProbShort] ? Model1(ModelInput1)
    AdjustedSignal ? AbstractSignalFilter(RawSignal, ProbLong, ProbShort)

    Update performance history from closed trades
    Compute AbstractFeatureSet2 from AbstractMetric_A and performance statistics

    ModelInput2 ? AbstractFeatureSet2
    [StopMultiplier, TakeProfitMultiplier] ? Model2(ModelInput2)
    Constrain StopMultiplier, TakeProfitMultiplier to allowable ranges

    StopDistance ? AbstractVolatilityMeasure * StopMultiplier
    TakeProfitDistance ? AbstractVolatilityMeasure * TakeProfitMultiplier

    If in training mode:
        Execute trades based on RawSignal
    Else:
        Execute trades based on AdjustedSignal

    Plot monitoring metrics (non-trading)