The ability to implement multiple machine learning models from both open source and proprietary software libraries. I like the fact that it provides a robust platform for doing many things in one place and easily allows you to modify or adjust how your models or data works. It could be more intuitive where people new into making models would find it easier to learn how to use the toolset. There were times when we were using encog and needed to perform tasks outside the scope of the program and there wasnt enough documentation and tutorials online to help us with those tasks. We used encoq to develop a predictive engine which helped our organization reduce the amount of manual labor required to provide customer service.