By Victor Thu
In order to avoid being disrupted or becoming obsolete in the near future, businesses are rushing to adopt digital business strategies.
According to Tom Koulopoulous, author and founder of the Boston-MA, based think tank, Delphi Group, devices increase by one order of magnitude every decade. Understanding what our future hyperconnected world will look like is challenging because we lack the shared experiences and behaviors to discuss it. Expectations are increasing and executives are consistently retooling their expectation sets to keep up with consumer demand.
Many have created new roles such as chief digital officer (CDO), to put together a digital enterprise plan. However, just as soon as they are building up their competencies around being a digital enterprise, they immediate realize their demand on IT has become exponentially more complex to support their digital transformation.
As a result, we are beginning to see the emergence of the chief automation officer, CAO.
The Role of a CAO
So just what does a CAO do? We see the CAO’s first priority to help sort out how artificial intelligence (AI) and automation will impact IT operations.
How does the CAO gain a full picture of what is happening to its environment?
In the last few years, three popular techniques have come to the forefront to address the ever increasing complex IT environment, adding advanced analytics to existing tools, robotic process automation, and general purpose AI.
However, all the above approaches have problem, which we detail below.
Adding deep analytics to existing tools is where the software vendor adds advanced analytics to the existing IT tools used today. Some have even branded that as adding machine learning and AI to these tools.
The challenge with this method is that these tools are only available in solving specific narrow challenges. For example, it might only be able to look specifically at networking issues versus the overall health of the IT environment. This is exactly the silo problems IT has today and going back to that will not help IT advance.
The issue with process automation is that there are a new crop of companies building robotic automation to simplify mundane tasks performed in IT. You simply program them to perform specific tasks.
What people are realizing is that robotic process automation technologies are literally robotic. These are highly deterministic programs that require engineers to create various scripts, or detailed instructions, directing robots on what to capture and interpret within existing applications.
This is intrinsically unscalable when you have hundreds of such scripts as they cannot adapt to the frequent changes in the IT environment. You frequently hear how companies need to hire a large number of staff to manage these automation scripts. That’s a very ironic value proposition.
The challenge with general purpose AI is that many companies have decided to use publicly available AI algorithms to solve their IT operational challenges. This is a great strategy for organizations that have resources to build a large team of data scientists and are empowered to tackle these challenges for the company.
The objection to this method is that this is a very long process. Companies need to spend time to build models to train their system.
Overcoming Challenges
Now that we know the challenges with the three approaches, what should we do? Is there a different way? The next generation IT solution can no longer sit in its own silo. That is the problem the entire industry has been suffering in the last couple of decades. We have done it and failed miserably.
It is therefore critical to connect business context and logic to an intelligent automation solution that can cut across the entire IT landscape. It needs to deliver meaningful business results as well as transform internal processes. CAOs are being put in place to enable a seamless digital transformation by overseeing the complex challenges IT operations.
Victor Thu is the global head of marketing at Digitate.
