Invenger Technologies, Inc

The Client

A leading US based Finance company.

The Challenge

Business was looking for methods to drive more Agents to the newly created dashboard application, to check the status of their claims online rather than calling to the HP status line. This will ultimately reduce the call volume and result in cost savings. A Web based monthly report should be generated, that identifies agents and their staff by user ID, who access the pages from the Dashboard.

The Issue

The Siebel Web Engines, at which the actual user requests are processed, are handling massive volumes of data. To track the user activity, we need to update the database on each page visit event, but this will add lot of overhead to the already busy servers. This can also result in poor response time. Besides this, on the report side, the expected number of records for 2 years is 6.4 billion (table size reaching up to 90GB). So the use of highly sophisticated techniques is indispensable.

The Solution

A combination of Siebel and .NET technologies is used as the solution. The user information is collected at the Business Component level in Siebel. This information will be stored as XML files, in a staging area. A console application [which runs as a scheduled task] scans the staging area for XML files and transfers the data to database for report generation. The processing is done on the basis of file creation time, in a FIFO manner, to reduce backlogging. The component also takes care of purging operation, after file processing. It is necessary to execute multiple instances of the application in parallel, to improve performance. The console application accepts command line parameters for required scaling purpose. Finally, a separate web application is used for the presentation of the reports.

To handle the worst case scenario of performance issue during peak business hours, a switch - like an LOV parameter - is implemented in the Siebel side, to turn off this processing completely, if found necessary.

Key Benefits

The status call volume was reduced by 5% via use of this report, which resulted in the reduction of approximately 56,000 calls per year. The potential saving per year is $3,20,000. There is also an intangible benefit for this solution. Once the data has been identified, it will be possible to develop website usage promotions to encourage online claims view for status.