MammothDB is a parallel, massively distributed analytics database. We are a hybrid architecture, running currently on stock Apache Hadoop, but we have our own Map/Reduce, and we run columnar databases on each node. We are fully SQL compliant, and shard the SQL across all the nodes using our Interactive M/R. MammothDB is a radically inexpensive alternative to traditional data warehousing (DWH) and business intelligence (BI) solutions.
For more information, visit http://www.mammothdb.com