ParAccel Extends Analytics to Unstructured Data

ParAccel, Inc., an analytic platform provider, announced the latest version of its flagship technology which introduces analytic integration for a wide range of analytic tools and data sources. ParAccel Analytic Database 3.1 (PADB 3.1) enables organizations to combine all internal and external data sources, both structured and unstructured, into analysis work.

"Analytic integration delivers new applications and analytic capabilities to our customers,"  says Tarun Loomba, chief marketing officer of ParAccel, tells 5 Minute Briefing. "Enterprises are accumulating more data than ever before. However, analysts are faced with two broad challenges - data is locked away in silos within the enterprise limiting the possibility of sharing data; and outdated technology has forced analysts to work with aggregated or sample data, as opposed to transactional level data, resulting in incomplete analyses."

PADB 3.1 includes high performance parallel connectors for legacy databases and Hadoop along with a flexible framework for integrating any proprietary or standard data source. "These ParAccel connectors leverage PADB's extensible analytics framework," Loomba explains. "And, they can reach out to these external data sources in a completely parallel fashion allowing bi-directional movement of data and analytics at very high speed."

Enhancements to the Omne Optimizer, Storage and Networking Subsystems are also included in the new version of PADB. ParAccel also introduced enhanced disaster recovery capabilities as part of the offering.

Loomba predicts that enterprises will soon be working more with unstructured than structured data files, and ParAccel is positioning PADB as a platform to address this shift. "Currently we are still seeing more structured rather than unstructured data," he says. "But the model we see becoming increasingly prevalent is the combination of structured, semi-structured and unstructured data in a single analysis. For instance in the financial services industry, the ability to analyze a portfolio and make trading decisions requires the base portfolio's historic information, based on structured data, which could be billions of rows, and new data feeds consisting of daily trade and quote data, which is semi-structured. Then, perhaps, per stock and ticker sentiment analysis, which would consist of  unstructured data from social media or blogs."

For more information, go to the ParAccel website