What statistical tests have you done to back up the claim "the ranking engine has been trained on a large amount of data and has high predictive power"? I'm ha

What statistical tests have you done to back up the claim "the ranking engine has been trained on a large amount of data and has high predictive power"? I'm having trouble connecting predictive AI models on highly biased portfolios only optimized on historical data to live high risk-adjusted returns and diversified portfolios. Do you have any numbers for the volatility of your portfolios and/or downside risks?