AutoViML was designed for building High Performance Interpretable Models with the fewest variables needed. The "V" in AutoViML stands for Variant because it tries multiple models with multiple features to find you the best performing model for your dataset.