Article

Data Update: Fueling our Global Endpoints

Data update: Increased automation and quality enables 32M new links to fuel our US and EU endpoints

For nearly thirty releases, we’ve scaled and optimized our Identity Graph to provide our customers with clean and up-to-date data. Nothing new here. But Data Version 29 marks a wave of firsts in our data story. This is the first release we’ve published since we’ve rebranded to Ekata. The first release that took advantage of increased automation to process more provider updates than ever before. The first release to pass our newly improved quality checks. And, most importantly, Data Version 29 is the first release to refresh our European endpoints.

Notable updates to our Identity Graph:

  • 22M new US person-address links, increased by 2%
  • 9M new US person-telephone links, increased by 3%
  • Refreshed person-address and person-telephone links in our top European nations: the United Kingdom, France, and Germany – increased EU online-offline coverage by 1%
  • Removal of outdated US business data to offer increasingly accurate matches for business search, decreased by 1%

As Ekata continues to grow as the international leader in identity verification, we are dedicated to offering low-latency solutions where our customers around the world need it. Given the volume of global sources that now feeds our Graph, it’s paramount that we reduce the cycle time of builds. This means less reliance on the human element, so over the past few months, we’ve engineered more automation into data builds. As a result, we experienced less variability between runs and were able to process well over a dozen updates in nearly half the time of past data versions– this means more global data more consistently.

However, despite the high volume of data running through our pipes, we never lose sight of quality. After all, what good is having a lot of data if some of it is garbage? As Ekata also continues to lead the fight against fraud, the quality of our Identity Graph becomes increasingly critical to fuel new machine learning models and best-in-class APIs around the world.

“Because of our commitment to quality as well as the advent of our Frankfurt, Germany datacenter last month, we’ve taken a step back to scale our processes to supply more global endpoints. By growing our analysis team, we’ve been able to expand our highly-specialized QA tools to simulate the impact of new entities before a build begins, and we’re now measuring the benefit of various sources to intelligently shape our release scope. In this release, we’ve cut back on data that no longer meets our high-bar for quality and added even more new data with increased confidence.”

–Justin Anderson, Product Manager, Data Services

Learn more about our advanced infrastructure and engineering capabilities to understand how we deliver unparalleled coverage, accuracy, and performance.

Thanks for reading! You might be interested in these posts, too: