SciDB is an open source database that organizes data in n-dimensional arrays. Interesting SciDB features include parallel processing, distributed storage, ACID transactions, and parallel time series and linear algebra operations. SciDB powers a number of big data projects including the NIH 1000 genomes browser. http://scidb.org.
Redis is an open source, fast networked key/value database that supports interesting structured values like hash tables, queues, sorted sets and more. Redis is incredibly versatile and popular. http://redis.io.
Dplyr extends the highly regarded split/apply/combine idea of plyr to data frames and related objects in external databases.
Bryan Lewis is the maintainer of the scidb and rredis packages, and is starting to work on integration between SciDB and dplyr. He'll show examples of all three projects. Bryan is a mathematician and chief data scientist for SciDB.