Vendors in the overcrowded data management space are becoming more and more creative in their efforts to stand out from the pack. The latest example of this increasing resourcefulness comes from Infobright Inc., which is rolling out a new offering today that promises to improve the performance of popular analytics systems such as Apache Spark and Teradata Database by up to a thousandfold.

Infobright Approximate Query, or IAQ for short, is based on the feature of the same name that the company added to its flagship columnar store a few quarters ago. Every time new records are loaded into an organization’s analytics environment, the software extracts a small sample and runs its contents through a model to try and mathematically make out the remaining information. The calculation produces what is essentially a sophisticated guestimate that can serve as a stand-in for the main dataset.

IAQ thereby removes the need for a query to be individually applied to every record in the system that it’s supporting. The functionality can come immensely useful for the large organizations that Infobright is targeting, which often have upwards of billions of data points to process. But there’s a major catch: An estimate produced by the system is bound to have at least some inaccuracies when such large amounts of information are involved.

The inherent uncertainty in the process limits the usefulness of IAQ to fairly high-level analyses that can make do with a small margin of error. Thankfully for Infobright, much of the work being performed by business intelligence professionals nowadays falls into that category, including everything from customer segmentation to measuring trends in sales performance.  The company is demonstrating some of the main use cases for its software at the Mobile World Congress in Barcelona this week.

Image via Geralt