Posted by: Steve Baunach

“House of Cards,” Netflix’s new hit political drama series, is not just good TV—it’s good TV because it was informed by Big Data. Netflix analyzed the viewing patterns of its 33 million subscribers to arrive at the following conclusions:

• Many viewers already liked the original BBC version of the series
• Movies featuring Kevin Spacey garnered a lot of views
• The work of director David Fincher achieved high start-to-finish stream rates

The result? “House of Cards” is now the most streamed piece of content in the United States and 40 other countries, according to Netflix.

Talk about a good business decision.

Netflix was able to analyze and cross-reference exactly what subscribers were watching, when, and on what device. They were also able to see when viewers would pause, fast-forward, rewind, and/or completely abandon watching a show. All of this information allowed Netflix to produce a show that viewers would not only like, but would also prove to be a solid investment for Netflix’s bottom line.

What does Netflix have to do with the Starview Enterprise Platform and Operational Intelligence, you ask? Here at Starview, we’re taking data in motion about our customers’ operations (as opposed to static data, like, the number of viewers that abandoned any given show) and similarly evaluating that data for patterns, correlations, similarities, and differences.

But, there is the crucial point of differentiation. Whereas Netflix execs evaluated their aggregated, static data over a period time to reach the decision to purchase “House of Cards,” the Starview Enterprise Platform allows users to build a decision making step right in to the analysis process, and trigger an action based on that decision, in real time, without disrupting the flow of other business processes.

To put this into a “House of Cards” perspective, it would be like if viewer input could change the course of the show while it was airing. This type of entertainment consumption may seem far off; but it’s exactly what the Starview Enterprise platform allows businesses to do, making a significant difference in industries like manufacturing, utilities, financial services, and telecommunications. These businesses especially need to be able to respond in real time and take action at the point of maximum influence in the business process, instead of waiting for static analysis while the opportunity at hand passes by.

While it used to be that there was no such thing as a sure thing in the entertainment business, data informed decisions are making that possibility more of a reality, and “House of Cards” is just the beginning. Operational Intelligence works in similar fashion by gathering, analyzing, and acting upon massive amounts of data while it is in motion for more efficient and timely decisions affecting your business’s bottom line. Think, how can your business benefit?

Posted by: Steve Baunach

New York Times Technology Reporter Steve Lohr recently penned a fascinating look at how humanities studies, from literature to history, are being advanced by machine learning and algorithms adopted from the Big Data technology space. One of the studies he reports used statistical aggregation to determine which authors in the literary canon had the greatest influence on other authors. Analysis of writing style, themes and word use revealed that Jane Austen—not Charles Dickens or Mark Twain—actually had greater influence over other authors. Another study evaluated the most “quotable” movie quotes from enthusiast websites and found that many fans’ favorite lines consisted of unusual word choices embedded within sentences of ordinary structure—such as “I love the smell of napalm in the morning” from the film “Apocalypse Now.”

While studies like the ones mentioned above are exciting for a number of reasons—especially for those interested in the marriage of qualitative and quantitative analytics—I think they’re most interesting for the commonalities to what we do here at Starview. The only difference is that we’re taking data in motion instead of static data (such a book already written, or a movie already made) and evaluating it for patterns, similarities, differences, and in some cases, egregious anomalies that might otherwise be missed due to the sheer enormity of information accumulating at any given point within our customers’ operations. The Starview Enterprise Platform also allows users to build decision-making capabilities into that analysis, and, if needed, to trigger an action based on that decision.

Imagine for a minute that every published author throughout history were alive today, and continuously adding to their respective oeuvres. Now imagine that as every new phrase is put to paper, simultaneously each—millions of words at once—is evaluated against all the other phrases also being created simultaneously. Imagine that any red herrings that arrive must be identified as quickly as possible and either deleted or changed, all without disturbing the flow of other phrases still accumulating. Otherwise the book, or manufacturing process, or financial transaction, might never be complete.

You still with me? As unlikely a scenario as this might seem, it’s exactly what the Starview Enterprise Platform is capable of. The technology of our platform can be used in just about any scenario where rapid analysis and decision-making on data in motion is required. Stay tuned.

Posted by: Steve Baunach

Wherever you stand on the political spectrum, it’d difficult be to say that Big Data did not play a pivotal role in the nation’s election just under a month ago. From The New York Times’ Nate Silver’s spot-on prediction of how the electoral votes would play out (based on deep analysis into information and trends), to the Obama campaign’s sustained use of information analysis to stay on the edge of trending topics among voters, the real winner this election year is data—mined, evaluated and acted upon.

Operational Intelligence works in similar fashion: massive amounts of data is gathered, analyzed and acted upon, all in real time. Obama’s “ground campaign”, which was adjusted according to constantly streaming feeds of information tapping into public sentiment and changing news events, worked according to the same principles as those that guide the way Big Data is used for Operational Intelligence in highly-sensitive, enterprise situations.

Whether you’re in the oil and gas industry and you’re looking into the data delivered from an immense network of pipelines and transportation hubs in order to determine the best way to quickly fix delays caused by unforeseen problem at one of your oil refineries; or you’re a candy maker looking to ensure not one piece of chocolate in your massive production facility gets wasted due to machinery malfunction, aggregating, analyzing and acting upon huge streams of disparate data as it occurs is more important than ever.

Posted by: Steve Baunach

The devastation wrought on the northeast by Hurricane Sandy last week is still being tallied. As awful as the impact has been on the millions of people affected, the lives lost, the homes destroyed, and travel plans disrupted, it’s becoming clear how much worse the situation could have been. While it’s hard to imagine a storm stronger than Sandy, amazing feats in technology, from incredibly accurate models to smart grid technology, enabled officials to keep citizens informed, to act proactively and close public transportation networks and other utilities, and to evacuate people in low-lying areas. Technological advances in Operational Intelligence will also speed recovery efforts — estimated to cost upwards of $50 billion.

What does all this have to do with Starview? A lot, actually. The prediction model used to anticipate the direction, impact and effects of Hurricane Sandy is not at all unlike the kinds of models that can be created and run to test a range of scenarios using Starview’s Active Analytics Platform. As variables in the real world change (a storm picks up speed, changes direction, etc.), then those same variables would instantly be adjusted by the Active Model—an essential component of our platform—to update the prediction model according to the latest conditions. The difference is that hurricanes move relatively slowly in comparison to the events that the Active Model can be used to predict, respond to and adjust for in the enterprise as they occur. Imagine using the same capability of prediction models in your enterprise to respond to rapidly changing events such as customer usage patterns / network traffic versus quality of service and then doing something about it in real time.

While most of us have never seen a storm of such epic proportions as Sandy, climate change models say that this hurricane most certainly will not be the last of such epic proportions in our lifetime. Regardless of where you stand on climate change, these models are worth examining. Besides the information they provide, they are also analogous to the kind of insight that Active Analytics and other associated technologies (such as Raytheon’s anomaly detection software) can provide.

Plus, thanks to the machine-learning and automated decision-making capabilities present in Active Analytics, these processes will not only become more efficient over time, but can be adjusted and improved in real time. How’s that for silver lining?

Posted by: Steve Baunach

Back in August, Harvard researchers reported that they had coded the entire contents of a book—including images, text and charts—into a genome, saved it, and successfully uncoded it back to its original format. Basically, what was once able to be saved on a floppy disk can now be saved on a single strand of DNA. Those terabyte portable drives? The data contained within one—entire libraries of information—could one day be stored in a test tube, or a petri dish. This is exciting on a number of levels. Just in the last few months, huge strides have been made in the ways organizations are utilizing Big Data—and it’s not limited to breakthroughs in storage. Processing, analyzing and acting upon massive amounts data in motion is just as big a challenge—if not a bigger challenge—than storing Big Data. A lot of the breakthroughs that have to do with handling Big Data in motion (another way of describing Operational Intelligence)are coming from areas that, like the DNA experiment, aren’t initially what come to mind.

Here are a few:

• Soccer: Adidas has introduced the miCoach, which monitors a player’s vitals, as well as his or her statistics in action, and instantly reports the information back to coaches. This is Big Data, in real time, on real human beings—imagine the implications if we all were fit with such monitors.
• Thermostats: “Smart” thermostats by Nest Labs are being used to help people save energy and more efficiently heat and cool their homes.
• Police Departments: The NYPD other urban police departments are using Big Data to map and cross-reference crime patterns around places, events, and other variables. I anticipate that taking this information and making it actionable in real time can’t be far behind—allocating law enforcement resources based on hot spots, time of day, and frequency of calls; tracking efficacy in response times; and measuring the best (and worst) possible outcome for crisis situations/in advance. This is Operational Intelligence at work in the world.
• What’s next? How else can Big Data be harnessed through Operational Intelligence to make the world a better place? We’re working on the enterprise side…what’s happening on your side?