One of the hardest applications to get right in the lexicon of enterprise applications is business intelligence.

As a segment, business intelligence applications probably chalk up more failures than any other application segment largely due to three reasons.

The first culprit is the simple fact that the IT people building and deploying these applications don't have enough knowledge about the business to create a truly effective application. Conversely, the business people that desperately need these applications don't know anything about how data is structured, which is the way IT people primarily think about any application. The end result is a general failure to communicate.

The second biggest problem is the very people that step into this void to try and help solve the problem. Consultants have turned business intelligence into a multi-billion dollar cottage industry. Unfortunately, the results are often spotty at best no matter how good the intentions are. The end result is all too often a project that is way over cost with a return on investment that might justify itself around the turn of the decade.

Finally, there's the nature of business itself. The simple fact is that we live in a world where applications need to respond to business requirements that are changing in a daily basis. All too often, the business intelligence application is essentially tied to a specific set of data that might only be partially relevant 30 days after the application is deployed.

One vendor that is trying to tackle all there of these issues head on is called Kalido, which has built something called an information management engine that stores data independent of the rules that used to construct a business intelligence application. That means that as requirements change, it's a whole lot easier to change or enhance any application built on top of the Kalido engine.

Essentially, what Kalido is trying to do is divorce the data that that represents the business from the applications that make use of that data. This approach is intended to significantly reduce the sense of being locked into the constructs of a specific application. Isolating data from the applications that make use of that data has been one of the industry's cherished goals ever since people first started talking about XML. But making that happen in practice has for one reason or another proved elusive.

Kalido may not prove to be the ultimate answer for everybody when it comes to business intelligence, but conceptually it represents a major step in the right direction when it comes to breaking the tyranny of data ownership that the all too often marks enterprise computing today.