IBM

Fusing real-time telemetry data captured court-side during the US Open with engineered base audio, the “US Open Sessions” website gave each user a chance to experience a tennis match as never before.

Each match was composed live, complemented with a WebGL-powered graphic visualizer, archived in a gallery and uploaded straight to SoundCloud. The sounds of the match were housed in an HTML5, JavaScript, and CSS3 website viewable on mobile, tablet and desktop.

For this project I created multiple server-side setups that were replicated in production across three clouds around the globe, each consisting of over 60 plus servers based on installation manifests that I wrote.

Each cloud contained servers built for specific tasks such as webservers, load balancers, processing servers and parsers (parsers ingested the live court data throughout the course of the event). Server installations included CentOS, Ubuntu, Nginx, Node.js, MongoDB, Redis, ImageMagick, Lua, Express, Forever and Python, plus a host of supplemental Node.js and CentOS packages.