The industry buzzwords “autonomics,” “cognitive computing,” “predictive data analysis,” and “automation” have been swirling around the industry for some time, but they are now moving full-speed from concept into practice. Those four terms – no matter how delivered – are the path to our future.
We are entering a new era in technology and IT is charged with implementing a new Knowledge Model to deliver the reality of a fully digital enterprise. A lot of technologies we’ve used are still in play, but the connective tissue becomes the threshold for success. The connective tissue is the rich set of data generated about our clients’ end users and their end points (both servers and desktops), and the associated data we have around our approach to supporting those devices and resolving faults.
The problem is that there are so many tools, technologies and approaches to automation – and implementation can be so complex – that in the rush to the digital enterprise, IT can potentially break more than it fixes. This can have the unintended effect of eroding the very service experience that we seek to enhance and deliver.
Circumventing this challenge requires a model that traverses service and configuration knowledge, providing information about end points that we want to support by leveraging these advanced technologies. The Knowledge Model has two important components, both Detection and Action:
- To not only detect an event or fault at an end point but also to understand the heuristics – the configuration, the operating environment and the other objects associated with it.
- To act, based on what faults took place at the end points and what programmatic actions should be taken to resolve the problem.
- The science occurs at the intersection of knowing what an object looks like and what fault took place at the end points.
There are many tools ─ scripting, orchestration, automation ─ that serve as mechanisms to deliver programmatic execution of a service. The challenge for IT is linking both sides of this Knowledge Model. The challenge for the industry is to avoid breaking more than what is fixed in the digital automation process.
We have effective models and mechanisms to integrate both end-point data and actions once we collect, classify and cluster data associated with the necessary actions. There is a learning curve, moving from supervised learning to unsupervised learning, but we don’t have to break more than we fix. We have to get this right, though, and autonomics, cognitive computing, predictive analytics and automation are the ways we maximize the data. The Knowledge Model shows us how to bring it all together and it’s an imperative: data is the currency of the future.
I’ll explore in future posts how to apply the Knowledge Model on the way to an automated, digital enterprise.