Published on 07-Jan-2013
"When we talk about the future with IBM PureData Systems technology, we're talking about leaving behind an old model where there's one system where you do operations and you work through things, and then there's another system where you can go and ask questions and report on other things. When you integrate those, there's a lot of powerful opportunity there, and the breadth of creativity is not limited." - Christine Twiford, manager of technology solutions and network engineering, T-Mobile
PureData System for Analytics (powered by Netezza technology), Big Data, Data Warehouse
With a customer base of 200 million mobile customers, T-Mobile saw its data growth rate multiply by three in just a year. Needing an integrated and agile approach to manage their Big Data, T-Mobile adopted the IBM PureData System to achieve unprecedented increases in analytic speed and accuracy.
The volume and the types of data that the communications industry collects and analyzes has been evolving rapidly in recent years. T-Mobile USA, Inc. (T-Mobile) collects data from roughly 90 different network feeds, and noticed annual data growth of roughly 75 percent. The company sought to stay ahead of the information curve by investing in Oracle data-warehouse technology, but quickly realized the limitations of the solution. T-Mobile needed a data-warehousing solution that would prepare it for the future.
T-Mobile implemented an IBM PureData System for Analytics solution, powered by IBM Netezza technology. The solution provides the client with a faster, more efficient data platform with the additional storage capacity it needed for its business. The client uses the solution to manage the 2 petabytes (PB) of data in its data environment, including 150,000 extract, transform and load (ETL) processes every day. In addition, the client uses the PureData System for Analytics solution to apply behavior-predictive analytics to its data, allowing it to determine when new network nodes need to be added.
By implementing a data-warehouse solution based on IBM PureData System for Analytics technology, T-Mobile gained a fast, efficient data platform with the storage capacity for its enormous data loads. In fact, the client reports that the solution allowed it, for the first time, to access and query nationwide data that had been far too massive for its previous system to handle. As a result, the client became more sophisticated about the types of questions it can ask and what features and services it can add.
Christine Twiford goes into detail on how the solution's integrated IBM Netezza technology has given T-Mobile the ability to spend time formulating creative questions and focusing more attention on discovering new insight rather than wasting time on maintaining the system. The client views the PureData System for Analytics solution as critical to its continued growth and industry success. Looking forward, she anticipates reaping value enterprisewide, moving beyond network engineering. Essentially, the PureData System for Analytics solution introduces an opportunity to be creative and empowered in a conversation with the data itself. Overall, T-Mobile reports tremendous success with the system and anticipates the opportunities that the PureData System for Analytics solution will bring in terms of providing features, services and other enhancements that improve customers' overall experience.
Christine Twiford – Manager of Technology – T Mobile
T-Mobile USA has about 35,000 sites all over the United States, nationwide coverage. We’re in 229 markets with 4G covering more than 200 million people with our latest 4G HSPA plus.
We’re seeing a lot of changes in the types of data that we’re collecting. We have about 90 different feeds that we’re collecting that are just pertaining to the network. And then also what we’re seeing is about 3/4x annual growth rate just from the regular traffic.
Today, we’re about 2 petabytes of data. That’s 150,000 ELT jabs everyday, about 20 billion rows that we load everyday.
We were able to take a look at the types of usage, what the behaviors were, do some predictive analytics, looking at megabytes, handsets and what kind of trends that we should be able to expect, and then we could decide that we could offer unlimited data without diminishing the customer experience, without clogging up the network.
With PureData Systems, we’re talking about leaving behind an old model where there’s one system where you do your operations and you work through things and then there’s another system where you can go and ask questions and report on, you know, other things. When you integrate those, there’s a lot of powerful opportunity there and the breadth of creativity is not limited.
We couldn’t believe that, first of all, that we were able to query all of the data at once. We had never had access to nationwide data because it had been too massive to query before.
We’ve had a lot of success building out our network because we have access to this data. We’re using it in pretty unique ways. We’re using the predictive analytics. We’re able to figure out when new network nodes need to be added and plan for that in a much smarter way. Having that kind of accuracy and timeliness is critical.
With an integrated approach, we’re able to spend our time being creative about the types of questions that we want to ask, getting the data out, establishing systems, organization systems that support this kind of interaction because it’s new, it’s so fast.
When I think about the future of big data and what PureData Systems offer as an opportunity, I really think about just the opportunity to be creative and to be empowered in a conversation with the data.
Big Data is important in the way that it’s used. It’s big because it has a big impact.
Products and services used
IBM products and services that were used in this case study.