Part 2: How do we increase ITIL maturity levels with Smart Machines?

Allan Andersen, Director of Enterprise Solutions, IPsoft

I have previously argued that ITIL has difficulties keeping up with the new iterative, Agile development methodologies that are transforming IT departments. That is not the only problem ITIL faces — arguably an even bigger issue is “lack of success” from a maturity standpoint. According to Gartner, “Overall, IT vendor management maturity for the approximate 1,000 assessments conducted since September 2010 is at a Level 2.4 (repeatable)” out of a scale of five.1This is after 30 years of ITIL, millions of certified practitioners, hundreds of certified ITSM solutions, and countless consulting organizations with practices around ITIL. While there are numerous reasons for lack of success and suggestions for improvements, they all tend to go back to project or program management, executive sponsorship and ongoing measurement. This leads us back to long, 18-24 month projects that adopt “Waterfall Development thinking” as opposed to Agile and DevOps methods and philosophies that deliver faster results.

So, how do we help increase the maturity levels around ITIL? First, as I noted in the previous post, we need to make it less bloated. Second, we need to utilize various forms of advanced automation and cognitive technology to streamline the process and remove humans from most of the repetitive tasks.

Let me describe how such changes can be enabled in IT by first looking at an example in a completely different environment: car services.

Only a few years back, I would call a car service when I needed a ride and an operator would dispatch the driver as needed. Today, everything is fully automated and the only human I speak to is the driver, and even that may change soon. Uber and its rivals have fully automated the car service process with human interaction only needed for exceptions. Not only that, their service is integrated with GPS, automatic payment, rating systems, optimized scheduling, notification to family or friends etc. This is often referred to as “combinatorial innovation” where most of the individual pieces are already in place, but combining these with new approaches can deliver a significant jump in how the service can be delivered and consumed.

Counter-intuitively “combinatorial innovation” seems to be lacking in IT operations. Improvements generally focus on incremental improvements but do not bring into question a fundamental change in delivery model. Yet, IT operations seems to be a logical candidate for digitalization, adopting new approaches, methods and technologies. The building blocks for this change are advances in autonomics, algorithms, analytics and cognitive technologies that create “smart machines” which can continuously perform, evaluate and recommend which activities can be automated, improved or even removed.

What is really different is that these smart machines are in the driver’s seat enforcing optimized ITIL processes by closing tickets, change requests (RFCs) and updating CMDBs in machine time. Human engineers are only involved in the processes when needed because of a unique situation, required skills or compliance reasons. Let’s take a closer look at how operations will run tomorrow:

  • Extreme Autonomics. Today most companies use individual scripts or orchestrated runbooks to automate tasks. However, it is still a manual task to actually identify the opportunities for automation and to create these flows. Tomorrow, we will have what I call Extreme Autonomics, where smart machines learn by monitoring and recording human engineers at work. Over time, based on analyzing all the activities observed, the smart machines will suggest the best approach to handling these tasks. Human engineers will review those suggestions, put final touches on the automatically created automations and then activate them.
  • Cognitive Service Desk. Today, end-users call a helpdesk, open tickets and place requests via Service Catalogs. The systems people use are designed by IT for IT and it can be a daunting experience for some end users to navigate this foreign IT language in search of what they need. Tomorrow, cognitive service agents will help these users in a far more business friendly fashion. Users will use whichever channel they find most convenient to contact the cognitive agent and log their issue or request. Communicating in plain language, the cognitive agent will ask the user questions for clarification and verification as necessary. It will be like having your own personalized service desk agent saying: Aww… did you forget your password again?
  • Pervasive Analytics and Algorithms. Today’s reporting and analytical solutions provide answers to the questions we have today. Tomorrow’s analytical solutions will provide insight and recommendations based on questions we haven’t yet imagined. Analytics will find anomalies, create “backlogs” for human engineers and continuously evaluate performance of the system and of human engineers for improvements.

The smart machines will be Virtual Engineers, Cognitive Service Desk agents and Robo-Managers certified in ITIL to name a few. It will be a collaborative environment where “smart machines” and humans are working side by side. Machines solve repetitive tasks at machine speed as well as assist human engineers to resolve more complex problems and outages faster. Human engineers will focus on driving continual service improvements with new types of analytics, evaluating and fine-tuning standard operating procedures (SOPs) and implementing service level agreements (SLAs) for key services.

The tomorrow I’ve described is not far away. In fact, much of this is happening today at IPsoft. By introducing smart machines into the heart of ITIL processes, we could quickly elevate the maturity of ITIL and forever raise the performance of IT Operations to new levels of efficiency.

1Gartner, “ITScore Assessments Highlight the Urgency to Accelerate Vendor Management Program Maturity” Christopher Ambrose, May 5, 2015