FullContact Issued Patent on Core Technology Behind Their Customer 360 Platform

“Even with tens of millions of updates to the graph per day, in our model we can completely recompute the entire identity graph within hours while maintaining the utmost accuracy.” Says FullContact CTO, Scott Brave.

FullContact, Inc, the leading open platform for 360 degree customer insights, today announced the issuance by the US Patent Office of a patent relating to the computational mechanisms that combine data fragments to form their identity graph.

The patent for invention number 9,684,696 covers the logic, processes and matching methodologies used to connect entities within this graph. Its purpose is to optimize both breadth and accuracy without sacrificing any computational speed.

“This patent makes claims in connection with our unique approach to identity resolution at scale. Making sense in real-time out of billions of individual data points–that are error-prone and changing frequently–can be a daunting task. This technology has served as the basis for what we’ve built and evolved over the past several years to effectively tackle this challenge,” said Bart Lorang, CEO at FullContact.

The logic of matching and appending customer data has historically involved joining exact keys or characteristics, like name or email or phone number. When that type of process happens in a relational database, an inexact match either gets thrown out or generates a new row. As the relational system gets larger, deduplication, updates, and cleansing jobs become more and more strained to the point that generalized optimizations become impossible.

Recognizing the futility of growing a traditional relational scheme, FullContact approaches the identity problem through a graph-centric data architecture that doesn’t force records into independent rows. Data is deconstructed, analyzed, and then compiled into interdependent nodes and edges in a probabilistic graph where the strength of those relationships and the quality of those characteristics are determined through real-world observations. As new data enters the system they are assessed in real-time using a layer of intelligence that gets smarter and more accurate as the graph increases in size.

“Even with tens of millions of updates to the graph per day, in our model we can…

Read the full article from the Source…

Leave a Reply

Your email address will not be published. Required fields are marked *