SQream Data Analytics Acceleration Platform
Entry Description
Due to exponentially growing data, organizations today are facing slowdowns, with analytics taking hours or even days. Time-consuming data preparation is needed for each change in perspective, and some complex analytics simply cannot be done.

SQream is a data analytics acceleration platform that accelerates analytics on massive datasets. SQream works as a speed layer alongside enterprises' MPP ecosystems, reducing long-running queries on terabytes to petabytes of data from days to hours and hours to minutes or even seconds. With SQream's GPU-accelerated processing power, enterprises can analyze a much larger portion of their historical and fresh data on more dimensions - for more accurate insights, better predictions, more reliable AI models, and better business decisions.

SQream analyzes raw data directly, enabling data scientists and BI analysts to freely explore their data from a variety of perspectives and dimensions, without the need for arduous preparation.

Leading organizations in telecom, retail, healthcare, finance, and additional industries around the world use SQream to improve quality of service, accelerate medical research, improve AI models, and gain access to previously inaccessible business insights.

Case Study
LG Uplus Cuts Analytics from Hours to Minutes, Enhancing Service for Millions of Subscribers

LG U+, the mobile carrier owned by LG Corporation, launched 5G services in December 2018. LG U+ wanted to improve network operations and efficiencies, reduce costs and downtime, and offer better quality of service to customers. These are a few examples of the game-changing processes they wanted to implement:

• The LG U+ Customer Service team needed to analyze, understand, and predict customer usage from the massive data generated from millions of users. This process was taking hours and sometimes days due to extremely long-running queries. They needed the ability to rapidly “slice and dice” their data.

• Network Engineering needed to better understand and model the noise on a call using frequency analysis to classify good quality calls. The data was gathered from multiple sources, and manually curated in a process that could take hours for each model. They needed to be able to rapidly update their models for better decision making while the data was still relevant.

• The Security team used a machine learning-based operation for analyzing data from multiple security sources in order to detect and prevent cyber and DDoS attacks on the LG U+ network. The operation had two parts - the training of the models and the execution. During the training phase, it was critical to sample as much data as possible, which could not be done with their existing system.

LG U+ chose SQream as their data analytics acceleration platform for its ability to rapidly ingest, compress, and analyze massive amounts of data, while meeting key performance indicators. LG U+ was able to significantly increase the amount of data they analyzed, while reducing long-running queries from days to hours, and hour to minutes.
Supporting Documents