Clarvoynt database integration quickly brings multiple systems together into a single view.  Lists of names and addresses now appear in a single table, with pointers back to the originating system.
However, if a customer has purchased multiple products from the bank (a mortgage, a car loan, investment, deposits, etc.) their name and address will appear multiple times.  Perhaps the spelling may be slightly different, or one system uses a middle name or initial and the next system does not. 

Customer addresses present a similar set of challenges. Resolving these slight mismatches to produce a cleansed single view of your customer brings great value to an organization.  Marketing, risk exposure, regulatory reporting will each benefit from a cleansed, de-duplicated master list of customer names and addresses.

The CeeqIT team provides this service quickly to any large organization struggling to solve this challenge. Using Clarvoynt, the name and address info from multiple systems is loaded and indexed.  The index automatically gives us frequency weighting of any name or address element.  We then use our name matching framework to check for similarities and apply a score, which represents a measure of confidence that two customers are the same, or two addresses are the same.

The scoring is reviewed and the results fed into our Machine Learning engine to improve our scoring model. The deliverable is a list of clients where each name is followed by a small set of possible matches from other database systems.  The scoring results are presented beside each candidate match. The final list is manually reviewed by bank personnel. If approved, the final list can be accepted, whereby a unique global identifier assigned to those names can be used to ensure future reporting identifies these names as being the same.

Additionally, CeeqIT can provide a set of SQL 'UPDATE' scripts for the approved and validated name matches.  With these scripts each source systems can be modified to contain the same spelling, case, and punctuation across all systems.