Michael Driscoll
Co-Founder & CEO, Metamarkets

Digital media publishers sit astride torrents of data about their content, audiences and advertisers. Every day, the world’s two billion internet users visit millions of pages, and are exposed to more than one hundred billion online advertising impressions. Publishers of digital content need to crunch this data to predict who will visit tomorrow, and how to price the ads those visitors see.

In the past year, Metamarkets has developed a Big Data stack for the media markets, comprised of three pieces:

1. a stream-based aggregation and processing service, built on top of the Kafka messaging system
2. a distributed, in-memory OLAP store, hosted on Amazon’s web services, that can scan and aggregate over one billion rows per second
3. a real-time web console, served by node.js, with dynamic data visualizations powered by D3, the successor to Protovis.

   Our platform is in currently in production with three of the largest media publishers in Europe and United States.