Zinia automatically produced forecast based on blending machine learning techniques

achieving 97% accuracy against actuals, which translated into an improvement of about 20% net present value for the balances lent (in comparison to their previous modelling)