Mark Flaherty (MF): A large part of the North American market data was available in a data warehouse which also included some of the forecast and budget information. But the real challenge that they faced was how to relate that on a global basis because the drugs in rest of the world Europe, Asia Pacific, et cetera were being primarily sold through wholesalers and distributors, and to get a full picture they have had to federate a lot of that information. So they had multiple challenges.
First the formats in which they were getting the data was very disparate and had to be normalized and accessed. Some of that was semi-structured data and had to be transformed, matched and integrated. But also they needed to be able to access it in a very dynamic real time way.
So the data services platform in this case is taking data from a data warehouse or a physical data store and mashing that up with a variety of data sources and then presenting the same virtual view of data in multiple ways.
| #1 Ranking: Read how InetSoft was rated #1 for user adoption in G2's user survey-based index |
|
Read More |
One of the benefits of the data services platform is you can have a logical view but then you can present it in different upfront formats. So in this case it can be pushed to an Excel spreadsheet or a query but it can also be accessed as a daily sales report, which is standard practice today, and they will continue to do that. But now they are also able to enable a Web portal and mobile applications to pull up specific drugs by specific drilldowns by geography, et cetera., et cetera, and also push that data in a batch mode to a historical data mart for analysis.
The last example I will use here is one around operations management. Operations management can be very complex whether you are talking about manufacturing operations, whether you are talking about customer service logistics, et cetera. In this particular case, it's about network management, provisioning IT services, as well as, network services in a very large telco company.
What they were trying to do is adopt the best practices for operations management dashboards where they would get real time views with drill down capability with some indication of causality in terms of how all of their systems were performing. Now this company had significant investments already in all kinds of specialized network monitoring suites. Some were IT monitoring, some were managing the actual telecom networks.
There was information about the applications and the support personnel and calendars for upgrades and all kinds of stuff including some semi-structured and unstructured data and log files, Excel, etc. Again here the challenge was not building the front end of the dashboards which will continue in a traditional model, but how do you provision those front ends?