Information Management (Big Data)
PureData for Analytics
IBM PureData for Analytics, powered by Netezza technology, is a massively parallel computing platform for analysis of very large datasets (hundreds of terabytes) that responds to queries in 1/10th to 1/100th the time of the traditional star-schema Datawarehouse running on commodity hardware and open source software. IBM PureData for Analytics provides up to 5TB per hour load speed, is easy to administer and requires little or no on-going database tuning.
Star-schema datawarehouses require extensive data transformation (the 'T' in ETL) to match new dataset structures to existing Fact Table structures, Dimension table indexes, and data mart boundaries. PureData for Analytics requires no Fact Table, no Dimension table indexes and no data marts. Most clients load new data directly into the PureData for Analytics platform and use the power of the massively parallel architecture to perform minimal transformations, thus enabling clients to re-purpose their ETL platform.
HHG's Analytics practice helps clients expand their business intelligence analytics to new datasets, extend analytical functions beyond standard Map/Reduce to regression, clustering and classification algorithms to provide behavioral modeling.
PureData for Analytics provides more than 200 in-database analytics functions which supports distributed analysis. The analytics plug-in from Fuzzy Logix extends the distributed analytics capability to over 800 functions and further reduces the time required to respond to queries.
HHG's complex event processing consulting then enables our clients to put newly created business intelligence to work in real-time using IBM InfoSphere's Streams to apply business rules and models to real-time transactions.