Tethering the Cloud ?

While the principle of cloud-based services hold many benefits for the business and consumer alike, there are many examples of where the purists vision of cloud computing may be at the cost of business benefit - be these cost-saving, security, resilience or other.

While a pure-cloud vision would hold that business application, processing, data-storage, analysis and reporting are all held in the cloud, there may be very good business reasons why part of the overall solution may be better served by premises-based add-ons.

Take security as an obvious example. Would salesforce.com lose it's pure-cloud mantra if, say, they provided a local NTU (network termination unit) that customers could deploy on-site to hold their highly confidential customer-specific data ? What if increasing internet speeds meant that the overall performance of the application was unaffected by the fact that some of this less-sensitive data was held in the cloud, while other data was held on this NTU ? Would tethering the cloud to this NTU fly in the face of the cloud purists or would the business benefits override the philosophy of the purists ? VPN being what it is could ensure that staff with the right levels of security could still continue to work anywhere on the planet and still see the benefits of an application that took some of it's data from the cloud, and others from secure NTU's within the customers own data-network ?

The reason for asking is explore how these principles apply to the delivery of voice services from the cloud. Just as with traditional business applications, there is great merit in abandoning the premises-based hardware/software solutions in favour of voice applications from the cloud. Why have to continually upgrade your PBX to the latest IP-whateveritis just in order to take advantage of features that, if delivered from the cloud, would work on any voice network, old or new ?

As you'll appreciate, the problem with the delivery of voice services from the cloud, unlike traditional business applications, is that the delivery of calls is currently metered. We all pay a pence per minute (or cent-per-minute for my american friends) for routing calls from the cloud to the handset...regardless of whether the call cost is paid for by the caller, the subscriber, or ad-funded - the network operator somehow makes a few pennies for delivering this call.

What if we tethered the voice services cloud directly to the customers premise so that while the processing of this call was cloud-based, the delivery of the call was over private-wires linking the customers own voice network to the public telephone network from which the call originated ? Would the requirement to deploy a voice NTU fly in the face of the cloud purists, or would rational business sense prevail ?

Take any readily embraced IT philosophy de jour, and in it you'll see a religion. You'll get the fanatics, the purists, the fundamentalists, the conservatives and the liberals. For a religion to become mainstream, it needs to find a place that accommodates all beliefs, biases and dispositions. It needs to work for everyone. Isn't this the case with cloud-computing that in order to be adopted by the mainstream it needs to address the concerns that businesses may have, be these around security, costs, resilience or other barriers to large-scale adoption ?

Time will surely tell.