high performance, scalable, NoSQL, extensible platform for data management

 Supports row-oriented storage (where all columns of a row are stored
 together), column-oriented storage (where columns are stored in groups,
 allowing for more efficient access and storage of column subsets) and
 log-structured merge trees (LSM), for sustained throughput under random
 insert workloads.
 .
 Includes ACID transactions with standard isolation levels and durability at
 both checkpoint and fine-grained granularity.
 .
 Can be used as a simple key/value store, but also has a complete schema
 layer, including indices and projections.