Big-data analytics, automation and artificial intelligence (AI) potentially increase business and network complexity. On the other hand, enterprise-cloud virtualization potentially reduces business and network complexity.
The case for virtualized cloud is pretty clear.
Previously, Adam Saenger, vice president of Global Product Developments and Management at Level 3 Communications, spoke with Telco Transformation about the organizational trends involving enterprise cloud solutions. (See Level 3's Saenger: Enterprise Cloud Means Business.) In Part II of this Q&A, Saenger further explains how the needs and pain points of the virtualized enterprise cloud are connected with those of automated big data systems.
Telco Transformation: Where are we seeing fear as a sticking point in terms of moving to an enterprise cloud environment?
Adam Saenger: Whether it's the cloud, SDN or any new technology -- call it the "consumerization of the enterprise" or "consumerization of IT" -- there's a fear that complexity is going to be masked through automation, and that that automation is going to yield less human intervention when it's really the opposite. It causes human intervention from an alignment perspective; alignment of business principles, alignment of standards within an organization and allows machines to work with machines where that makes most sense, and humans to work with humans where that makes sense and humans to work with machines where that makes sense.
TT: To speak of automation, to what extent do you see the merger of big data and AI technologies coming in to play to help answer enterprise-cloud customers' needs?
AS: Data and AI and cloud and network connectivity, all of these things are coming together. I believe they need three things: visibility, control and security.
When those things come together, the customer is empowered to solve their challenges faster, and the challenges that they solve will be very different across industries whether it's healthcare, retail, manufacturing, IT. I'll use healthcare as an example. There is a wide, wide lens about the appropriateness of cloud in healthcare and a fear factor of how secure is the cloud infrastructure: "Do I lose control?"
Let's take a patient example. A patient goes in for treatment in a remote clinic, and this remote clinic may not see this person regularly. They need access to the person's healthcare records as quickly as possible, and the access into and transmission of those healthcare records needs to happen in a very quick timeframe.
Very quickly, this patient's care escalates to needing to do some remote diagnostics, to which that clinic can connect into hospital resources and get the right resources connected not only quickly, but securely, and to the level of performance that a video diagnostic will allow. That has taken big data -- the data that is available on that patient, mashed up against the data that has what that condition may yield the symptoms -- and the control, the SDN nature of being able to change the class of service allocations or the bandwidth allocations to that location so that the video transmission doesn't overtake the other day-to-day operating needs that clinic has. So that that patient gets the best care where, when and how they need it.
And then envelop the entire thing with security. If your financial information gets out, you get issued a new credit card pretty quickly, and the anomalous traffic that has been used or the data that's been used to identify that anomaly allows them to shut that down quickly. But, if your healthcare information gets compromised, that's your personal information, and getting new unique personal identifiers assigned happens very, very infrequently, if ever. And so that information has to be secured.
So I think those things all come together, and it's driven by the visibility into the service and into the data, the control of all aspects, and the security of the information that is being transmitted and/or accessed any given time.
TT: As automation and AI and the like gain ground in the enterprise and networking, to what extent must AI networks themselves rely upon enterprise cloud?
AS: I would say that AI relies on access to data. Whether that data is stored in a public or a private cloud, that data still needs to be accessed, learned and evolved. So I would say it is dependent upon cloud and, to a great extent, and even more so, it's dependent upon ubiquitous access to that information. The importance of the network increases over time. So I believe it will be a driver of both cloud and network consumption. We're living in a world right now where the Internet of Things is becoming reality, and these devices, the things that are being connected, are becoming more and more intelligent. The cloud is one mechanism that will be used to support, allow and further the mass adoption of IoT.
— Joe Stanganelli, Contributing Writer, Telco Transformation