By Salvatore Ulto, Consultant
The days of a one ‘mainframe-like’ database system that a company is dependent on to holds its mission critical data are long gone and as a result data duplication and data variances can exist within an organization. Data has now become more dynamic and company data is being updated by its employees, customers and third parties. This presents unique problems to organizations in helping them to understand and present their data in a manner so that they can make intelligent business decisions.
We at Bardess have a complete suite of data management solutions from data quality analysis to data cleansing and improvement to data revitalization and it’s important that we stay up-to-do date with the latest tools and technologies in order to meet our customers’ needs. One tool that has caught our eye is Tibco patterns which is a mathematical model that finds patterns of interest in data – helping people and systems find what they need.
In future blogs, I will put Tibco Patterns through a POC to see how it can help in terms of matching site information between two disparate systems and report back my findings. To begin the POC, let’s define the business problem.
A client is trying to marry information from its call center database against its contracts database. Customers call in for service on parts claiming that the parts are covered under contract. The client is looking for a report that shows the device that was reported and whether or not it is covered under contract. The contracts database keeps track of parts that are under contract for a particular site and what the SLA is for repairing/replacing the part. The call center operators are under their own SLA and there is no tie in between the site information that the customer provides and the site information that is in the contract database.
This presents the classic address information issue in that address elements in both the call center and contract databases are free form and matching information between both systems is manually intensive in nature. A report is produced monthly and normally takes about 16 man hours to complete, with the majority of the work being matching the site information from the call center against the contracts database.
Using raw data from a previous report, I will send the call center’s site information into Tibco Patterns to compare against the ‘master’ site information in the contracts database and see how it matches the information and at what degree of certainty. Lastly, I will compare those results against my manual results for that month to see how close they are and help gauge how much time will be saved using Tibco Patterns.
Stay tuned for my results . . .
About the Author
Sal Ulto is a senior consultant for Bardess Group Ltd working on project and resource planning. He has over 20 years experience in Project Management as well as Business Process Solutions. Sal comes to Bardess working with the Finance Industry where his last role was analyzing and reengineering business processes for the IT Asset Management lifecycle ranging from request through fulfillment.