Corporate leaders, at various managerial levels, have been challenged by fears that any day soon, a competitor of inferior technical capacity would jump ahead from behind. They were amazed at the fast and easy practice when we demonstrated the RapidMiner platform. Some of them have also suffered from ill-fated projects in the past, perfectly aware of the negative impact due to slow, error-prone, and delayed data preparation, model training, and deployment. Understanding its advantages, they too realized that others can easily outrun them quickly with RapidMiner.
Granted, companies are looking for vertical solutions aiming directly at their specific applications while RapidMiner, albeit powerful, is a platform. In addition to current communication by RapidMiner and the enthusiastic community, efforts must be spent by both sides, the customers and the solution providers like us, to practice, solve problems, and build up confidence so that gaps can be bridged.
It is rather idealistic to wish for a vertical solutions aiming directly at specific applications. You may have concerns that RapidMiner looks powerful but is still a few steps short. You are not alone but others have used RapidMiner to complete data science projects in weeks if not in days. With patience and determination to become a good problem solver, RapidMIner will bring you far ahead of competitors in a short period of time.
Data science is after all an art of problem solving. We have encountered problems such as, in predicting disgruntled customer, “can we generate some statistics for the past calls that belong to the same customer of the incoming call?” and, in predicting the lab-generated quality measure in a chemical plant, “can we shift each and every dependent variable according to their respective time lags?” Yes, RapidMiner can support them and thus incorporate the steps in a RapidMiner process. Such problems are however slightly beyond the basic practice of RapidMiner.