How Buzzfeed Uses Real-Time Machine Learning to Choose Their Viral Content

by on October 27, 2016

This talk took place at the Domino Data Science Pop-up in Los Angeles, CA on September 14, 2016.

In this presentation, Jane Kelly, Director of Data Products at Buzzfeed, talks about how the popular online media company uses real-time machine learning to inform what content to promote, before it goes viral.


Listen in to learn about the innovative platform Buzzfeed has developed to:

  • Capture data and feed it to a centralized hub for real-time analysis;
  • Support collaboration between engineers, data scientists and content creators;
  • Select the right features to identify posts that should be translated into other languages;
  • Balance rapid experimentation with product stability, and effectively publish their data to different consumers internally.

Slides for Jane’s presentation can be found here.