My thoughts on reading #3

For this week’s reading, the aspect that jumped out at me is the focus on information sharing and breaking down silos.  There are two reasons why this jumps out at me, 1) I have been hearing about this topic since I landed my first job in IT 20 some odd years ago and apparently we have not been able to solve this problem yet and 2) my company is investing heavily in trying to solve this same problem.

I partly see the reason this is still a problem for so many organizations (from past experience) is that to solve this problem, organizations look for a “magic app” or product that will solve this issue by simply buying something.  And I have to admit, that has been my approach to the problem too.  The reading has for me outlined the problem/solution in a more comprehensive way, framing it as an Enterprise Information Architecture (EIA) problem/solution.

There will always be the need to store data in silos (speed of processing, COTS product data schemas, Line of Business custom applications).  But taking this reality into a broad EIA approach provides the methodology to pull them all together into an integrated whole.  As stated in the reading, the key is defining enterprise data in the abstract first and then planning for information sharing and integration approaches.

My company is currently in the early stages of working thought this.  We began development of what we call the Enterprise Data Infrastructure.  From a Business Architecture standpoint, this includes definitions of what our Enterprise Data Objects are (Party, Security, Contract, etc.).  From a Data Architecture perspective, this includes defining separate logical data schemas and physical databases aligning with each of these entities.  From a Technology Architecture perspective, a Data Services layer provides the translation and distribution services to both coordinate data relationships between the different entities, but also with other systems.  Communications with other systems includes definitions of update events to be communicated out to other systems as well as incoming additions/updates/deletes from distributed systems of record.  Rolling this all out involves defining highest priority enties and systems first and enforcing an approach of “getting the data from EDI” whenever it becomes available, in coordination with existing applications’ release schedules

As this is the first instance of an actually concrete solution to this Enterprise level problem, I am curious if others from the class have encountered/developed alternative solutions?