Real-Time Analytics At Work Inside Telcos - InformationWeek

InformationWeek is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Data Management // Big Data Analytics

Real-Time Analytics At Work Inside Telcos

What can real-time analytics reveal? A mobile provider in New York City recently realized that 10% of its capacity was being consumed by a single application.

16 Top Big Data Analytics Platforms
16 Top Big Data Analytics Platforms
(Click image for larger view and slideshow.)

Much of the discussion on big data centers focuses on effective ways to store growing volumes of digital information. But, of course, size isn't everything.

"It's not just the fact that there's big data, it's what you do with it. If you have that insight and don't act on it, then it's wasted effort," IBM predictive analytics expert Richard Rodts told InformationWeek a year ago.

Indeed, one of the business world's concerns is exactly how to convert all those bits into workable strategies, the practicable intelligence required to justify an investment in big data. According to Anukool Lakhina, founder of big data analytics firm Guavus, real-time analytics can provide the catalyst needed to achieve that ROI.

"I hope to see more conversation around the value of big data versus the size of big data," said Lakhina in a phone interview with InformationWeek.

The Silicon Valley-based Guavus sells big data software that helps businesses, mostly telecom companies, analyze "live" data, an approach that promises faster insights than the traditional store-first, analyze-later method. It counts seven of the top 10 US cable operators, and three of the top four US mobile carriers, as its customers, said Lakhina, who in addition to founding Guavus eight years ago also serves as the company's executive VP of strategy.

[Looking for a big data job? Read Big-Data Hiring: 5 Facts From The Field.]

The technology behind Guavus's real-time analytics emerged from Lakhina's PhD work more than a decade ago. "It goes back to a science project that was done at Sprint Advanced Technology Labs," he told us. The project's main goal was to help Sprint take incoming data on its mobile subscribers and continuously analyze it, thereby enabling the carrier to make timelier decisions for a range of use cases, including network security, engineering, customer support, and marketing.

"Sprint had a very audacious idea 10 to 12 years ago to instrument their network with very sophisticated data probes and collection devices. The idea was that once they could collect all this data, they could get improved operational intelligence for difference business functions."

Not only did the project help Lakhina earn his PhD, it created the technical foundation for Guavus.

"We realized that the traditional approach of taking data and moving it to a central place to store it, and then ask questions about it, would not work. You had to take a different approach. The centralized, store-first model wasn't the right way to solve this problem. One had to take an analyze-first, distributed approach."

Real-time analysis is playing an increasingly important role in data-driven enterprises, particularly in the telecom market, said Lakhina.

"On the network engineering side, one of the challenges that all mobile operators have with the advent of the smartphone is that they're seeing unprecedented demands on their networks. It's very difficult for them to engineer and plan their networks to provide the optimal quality of service."

Analytics tools from startups like Guavus -- as well as much larger players like IBM -- give mobile carriers up-to-the-minute visibility on exactly what devices are consuming content on their networks, and where the congestion is.

Lakhina provided this example: A mobile provider in New York City recently realized that 10% of its capacity was being consumed by a single application.

"It turned out the application belonged to a taxi cab company," said Lakhina. "If you've been inside a New York taxi, you'll know they have these live video news streams. This company had taken the credit card transaction application and rejiggered it to carry live streams."

Customer service reps also benefit from real-time analytics, which enables them to quickly determine why mobile subscribers exceed their data plan limits, added Lakhina.

Engage with Oracle president Mark Hurd, NFL CIO Michelle McKenna-Doyle, General Motors CIO Randy Mott, Box founder Aaron Levie, UPMC CIO Dan Drawbaugh, GE Power CIO Jim Fowler, and other leaders of the Digital Business movement at the InformationWeek Conference and Elite 100 Awards Ceremony, to be held in conjunction with Interop in Las Vegas, March 31 to April 1, 2014. See the full agenda here.

Jeff Bertolucci is a technology journalist in Los Angeles who writes mostly for Kiplinger's Personal Finance, The Saturday Evening Post, and InformationWeek. View Full Bio

We welcome your comments on this topic on our social media channels, or [contact us directly] with questions about the site.
Comment  | 
Print  | 
More Insights
Newest First  |  Oldest First  |  Threaded View
SeanBowen PUSH Technology
SeanBowen PUSH Technology,
User Rank: Author
3/10/2014 | 5:25:30 AM
Re: Real-time Analytics at Work Inside Telco
This example is of course extreme, but it highlights one very real problem for mobile providers: With mobile data consumption growing at the rate it is, networks are more heavily congested than ever before. Operators are finding it increasingly difficult to provide the optimal quality of service to their users.

But, this is not the only issue that mobile providers are facing right now. Not only do they have to deal with the explosion in data consumption and growing bandwidth demand, but they also need to work out how to recover revenue from over-the-top (OTT) services, and how to make money from the cloud. They must consider how to build new business models that lessen the strain on the network by reducing the amount of data sent across it. They also need to consider how to climb the value stack by offering new cloud services.

What's needed is an intelligent solution that optimizes how data is sent over the network. Some telco providers have already turned to our real-time data distribution platform to gain back network control. Some of the world's leading telcos are choosing Diffusion to:

- Radically improve performance of real-time conversational applications and services

- Build smarter data services that reduce the amount of data sent across the network

- Significantly save bandwidth

- Reduction in hardware footprint and associated infrastructure

- Increase customer retention

- Develop new high performance cloud services

One example is our cooperation with Telefonica Digital. Combining Diffusion and Telefonica's infrastructure-as-a-service, offers the first high performance, scalable data distribution solution from the cloud to any device. This agreement creates an entirely new method of building scalable mobile and web services, and means customers can deliver real-time applications at low cost.
User Rank: Ninja
2/28/2014 | 8:44:13 PM
Digital Data
Organizations which create and store more transactional data in digital form will be able to collect more accurate and detailed performance information.
User Rank: Ninja
2/28/2014 | 8:39:07 PM
Re: Realtime vs. Analyze Later
@brain I agree with your comment. Real time data analysis will definitely help you to take effective decisions.
User Rank: Ninja
2/25/2014 | 1:35:28 PM
End Game
I'm wondering what came from realizing what the cab companies were doing? Did they, or could they, shut that traffic down? This is the Netflix dilemma in a nutshell. Does this carrier now go out and triple it's capacity, with all it's customers paying higher monthly charges, because this cab company wanted to create a TV service out of something that was never intended to be that?

Where does this all end?
D. Henschen
D. Henschen,
User Rank: Author
2/25/2014 | 1:19:11 PM
Event processing technology?
I'm not really clear on the technology in play here. Event processing and analysis of streaming data (while it's still streaming) has been around for a long time. How does Guavus compare with the event-processing technology offered by the likes of IBM, Oracle, SAP and others that has been used from everything from financial trading floors to network security to national intelligence (including the likes of the NSA)?

The in-memory database camp is also addressing real-time needs, with legacy products like Oracle TimesTen and IBM solidDB used by telcos and financial services to quickly store and analyze data without having to wait for overnight batch data loading. SAP Hana is not horning in on this space for more mainstream data warhousing and enterprise app needs. An there's a raft of NoSQL and NewSQL database players like Aerospike, VoltDB, MemSQL, Pivotal GemFire and others exploiting RAM to analyze data and trigger decisions within milliseconds.

Maybe Guavus can tell us more about where it fits in this mix?

User Rank: Strategist
2/25/2014 | 10:08:01 AM
Realtime vs. Analyze Later
Great article.  It appears to me that this need to analyze data in realtime falls under more of an operational business intelligence model (if there is such a term).  The need to store now and analyze later (i.e., looking at monthly trends at the end of a fiscal year) offers a higher-level view of data for making more long term decisions (not sure what the term for that is).  The former allows you to analyze a few trees quickly and the latter allows for a larger view of the forest.  Both are valuable but they each offer different kinds of value.  Agree?  Disagree? 

Anyway, this is a great, great article, thanks so much for sharing.
2021 Outlook: Tackling Cloud Transformation Choices
Joao-Pierre S. Ruth, Senior Writer,  1/4/2021
Enterprise IT Leaders Face Two Paths to AI
Jessica Davis, Senior Editor, Enterprise Apps,  12/23/2020
10 IT Trends to Watch for in 2021
Cynthia Harvey, Freelance Journalist, InformationWeek,  12/22/2020
White Papers
Register for InformationWeek Newsletters
Current Issue
2021 Top Enterprise IT Trends
We've identified the key trends that are poised to impact the IT landscape in 2021. Find out why they're important and how they will affect you.
Flash Poll