To help businesses maximize the value of their big data investment, Intel Corporation launched the Intel Data Platform, a software suite based on open source technologies designed to make it easier and faster for companies to move from big data to big discoveries.
Additionally, the new Intel® Data Platform: Analytics Toolkit (Intel® Data Platform AT) will create a graph analytics and predictive modeling environment to help businesses uncover valuable insights from hidden relationships within data.
The Intel Data Platform provides an open environment for import, management and analysis of big data that builds upon the Intel® Distribution for Apache Hadoop* to offer improved reliability and enterprise-class security and support. The platform also features several new data processing capabilities including streaming data processing, interactive and iterative analytics, and graph processing. Together these capabilities enable enterprises to extract value from data in ways not previously feasible with Apache Hadoop alone.
"As big data shifts from hype to reality, Intel is helping to break down the barriers to adoption by easing complexity and creating more value," said Boyd Davis, vice president and general manager of Intel's Datacenter Software Division. "Much like an operating system for big data processing, the Intel Data Platform supports a wide variety of applications while providing improved security, reliability and peace of mind to customers using open source software."
With fully integrated frameworks for stream processing in real time, companies such as retailers can use the Intel Data Platform to quickly analyze social media and sensor data, in-store purchases, and inventory to gauge the impact of a celebrity endorsement on the demand for a new product. Using the interactive and iterative frameworks for big data applications, industries such as telecommunications can look up a vast array of data tailored to a specific customer in order to personalize recommendations for new products or services.
The platform will be available next quarter in two versions – Enterprise Edition and Premium Edition – each offering various levels of support. The Enterprise Edition will offer full platform capabilities as a free software product to customers who can support their deployment. The Premium Edition will be available for purchase on an annual subscription basis and will provide additional technical features including enhanced automation, proactive security fixes and alerts, ongoing feature enhancements, and live telephone technical support.
Uncovering Surprising Answers to Unforeseen Questions
Today's tools for data analysis are often capable of answering the known questions but are blind to unexpected connections within the data. The Intel Data Platform AT is designed to reduce the complexity, effort and cost associated with graph analytics and predictive modeling to help data scientists discover knowledge and find hidden relationships within data.
The toolkit provides a foundation of common algorithms, such as graphs and network-based clustering, that IT teams can build on and customize with domain-specific code. The easy-to-deploy algorithms are broad enough to be applied to multiple industries, including financial services, healthcare and retail. The toolkit will also provide an enhanced development framework for unifying graph analytics and classical machine learning to ease the programming effort.
Using the toolkit, a data scientist at a financial services firm can gain richer insight by developing a fraud detection service that identifies patterns between purchasers, merchants and transactions to uncover potential points of compromise instead of only monitoring the purchase habits of individuals. The Intel Data Platform AT is available in beta now and expected to be commercially available by the end of the second quarter.
From Hype to Action: Big Data in the Real World
Intel is already working with companies of all sizes to help uncover value in data to enable new scientific discoveries, business models and consumer experiences. By implementing Intel-based hardware and software solutions, China Mobile Guangdong* was able to improve billing processes and customer service by enabling online bill payment as well as the retrieval of up to six months' worth of call data records in near real time. China Mobile Guangdong's detailed billing statement inquiry system can now retrieve 300,000 records per second and insert 800,000 records per second or the equivalent of 30 terabytes of subscriber billing data per month.
Intel also worked with Living Naturally*, a retail technology provider, to develop business analytics algorithms based on the Intel Distribution for Apache Hadoop to help retailers better manage supply chain and product promotions. The algorithms analyze a mix of internal and external data, such as social media, search engines and weather sites, to provide retailers with better insight and help determine when to reorder products in optimal quantities to minimize surpluses, shortages and shelf life expirations. For example, when a popular medical expert with a large Twitter* following recommended raspberry ketone pills to help reduce weight, the result was a spike in sales and empty shelves. The Intel-based analytics solution can help retailers anticipate increased sales demand based on social media or other online postings.
