In early 2012 I joined the teams working on the newest Cerner platform, Millennium+. It was an ambitious re-architecture of Cerner's electronic medical record platform to enable new types of applications.
I worked on application services that enabled semantic search of different areas of a patient's chart. This involved building the search indexes, writing batch and real-time processing jobs to index this data, and writing services to make the data available to applications.
The search indexes were built on Solr, and the services built with Java and Groovy using Apache Thrift for a protocol.