A question I receive quite a bit from people who are newly introduced to NoSQL platforms and the overall concept of multi-structured data sets is the following:
Can you use NoSQL for business intelligence and analytics?
This is a perfectly acceptable query. You have returned from a conference on NoSQL or big data and you are looking for ways to utilize this newfound domain within your organization. Or better yet, you have been sent to do research by your boss who has returned from one of the previously mentioned conferences or who found a “shiny new toy” article in a technical magazine.
However, I have often found that the question is better redefined as the following:
Can you use a NoSQL data store to provide data to a business intelligence platform or an analytics system?
In fact, a more direct version of the question might be:
Can I connect a Hadoop HDFS instance or MongoDB store to my BusinessObjects or MicroStrategy implementation and have it work without problems, because my boss wants it to?
To be perfectly honest, these are two different types of questions, and I think that professionals pushing the envelope of their business should understand the difference(s). I will try to put some context around the concept.