Codefry has implemented enterprise applications to improve business decisions across industries and professions. Our first solution discovers at astonishing speeds entities of interest within data allowing enterprises to see previously hidden relationships among data items. This is the first step towards enterprise analytics based on plain text data.

Entity Discovery applications allow users to find data items of interest within unstructured or insufficiently structured input. The have been used to successfully find the following data items: full address of a location, partial address, facility abbreviation (out of a list of 300,000) and world airports. In each case, the application properly geocoded items and added all of the necessary metadata in order to “enrich” the data. Enriching means inserting additional information about the item that can be used to match it to related items, searching, and performing comprehensive data analysis.

There are only a few systems performing the technically complex process of Entity Discovery and Enrichment with precision and speed. The challenge is the number of comparisons necessary to find a match. For a geographical address, each of the words in a phrase has to be compared against several million valid world addresses. Our system uses fast AI algorithms and efficient programming language (C++) to provide outstanding performance.

As a result, these data feeds can be processed in real time. Our system has been used to process historical data going back several years with large numbers of geographical references, as well as on live data messaging feeds with data rates of up to 200 text messages per second. In both cases, the system showed high quality as well as speed of operation.

Entity Enrichment solutions transform data into a valuable resource available for analytic and visualization tools. These data-intensive capabilities improve business decisions making and data insight across industries.