Gearing up for the Internet of things
In that example, Rishi is talking about customer-facing systems, otherwise known as systems of engagement. But he goes on to discuss what he calls cloud-based "systems of interaction," where sensors deliver data and events that core enterprise systems respond to -- a model similar to the event-driven agility Allison says enterprise CIOs want to instill in on-premises systems. In other words, Rishi is talking about a cloud platform for the Internet of things, custom-built by IBM (and deployed on IBM SoftLayer, among other clouds), while Allison sees his clients gravitating toward an on-premises version.
Guess what? Salesforce1 is primed to ingest such telemetry as well, as is Amazon's Kinesis service. The VMware spinoff Pivotal was conceived from the ground up with the Internet-of-things paradigm in mind.
Systems of interaction, as Rishi calls them, are crucial to the future of the cloud. Can the Salesforce multiplatform behemoth stretch to accommodate a huge influx of machine-to-machine data? You'd think that, as a first mover, Salesforce would know how to scale pretty well by now. On the other hand, as the first big multitenanted SaaS provider, it has operated continuously for its customers year after year as it has rolled out countless upgrades on the fly. Will architectural flaws be exposed in dramatic fashion when a whole new array of capabilities are piled on?
The truth is that the public cloud is not transparent, so it's almost impossible to know. Sure, many enterprises are happy to enhance the customer experience with new cloud-based systems of engagement. But as those systems get closer to the core, and events are processed that affect critical systems, intangible issues of trust and even Hawaiian shirts may sway the final decision.
This article, "Software as a service vs. old-school IT," originally appeared at InfoWorld.com. Read more of Eric Knorr's Modernizing IT blog. And for the latest business technology news, follow InfoWorld on Twitter.