Having a highly differentiated set of automation and AI capabilities is a doubled edged sword.
On the one hand it provides an exciting platform for growth as Arago helps clients to redefine digital operations through our decision-making engine AI platform HIRO. At the same time this allows them to accelerate their digital journey while fending off disruptors.
On the other hand, this differentiation doesn’t lend itself easily to be pigeonholed and tell the story by following a well-trodden path.
I was therefore delighted to hear our partner, SYSback, not only talking about their experience with Arago but also demonstrating HIRO at a recent AI conference in Cologne. Talking to Olaf Windhäuser, Senior Vice President Services, and his team at that conference provided valuable insights into the lessons learned from the early deployments of AI.
While we are going through the process of changing the way we position and communicate our capabilities, I took the opportunity to catch up with Olaf to follow on from our discussions in Cologne.
Olaf, I must admit I only came across SYSback once I joined Arago. Could you summarize what SYSback is all about and what you are trying to achieve?
SYSback is a 16-year-old IT company in Germany headquartered in Hamburg with locations in Darmstadt, Cologne, Dresden and Munich. We were founded as an IT company that sold hardware as well as software and made professional services projects with those components. I have been at SYSback for 4.5 years now and together with my colleagues on the SYSback board, I am turning the company from a reseller into a proper IT service provider.
Our business model today is called “Holistic Automation”. We automate everything end-to-end, meaning processes as well as technology. This is a big part within Digitization today, since data is everywhere and needs to be automated so that enterprises can apply them in new business models. We bring ITSM and Automation (Runbook, RPA and AI-based) together to generate end-to-end automation and to overcome all the silos that customers have typically built in their IT worlds over the last couple of years. Nevertheless, in times of Hybrid Cloud, silos have stopped clients from delivering services effectively and therefore need to be overcome to implement ITIL and DevOps at the same time.
In 2019 we will expand our ITSM and Automation capabilities into a Managed Service offering, which will give us the USP becoming the only Managed Service provider in the German market that is capable of delivering highly automated and therefore high-quality Managed Services to midsize companies as well as larger enterprises. The other differentiator is that our offering is delivered out of Germany from our hubs in Hamburg and Darmstadt, thus with perceptions of great quality, high SLAs, and still reasonable cost.
As an early adopter and proponent of Arago what attracted you to our approach and how do you tell the story to prospects?
The approach of knowledge building blocks that are captured and codified by HIRO is a unique approach in our portfolio. Our automation solutions typically follow the traditional runbook approach. HIRO also offers the option of reusing parts of the implemented use cases, i.e. the atomic knowledge building blocks, in other use cases, thus, reducing the development effort required in automation.
This is a great advantage, especially for customers who want to automate on a large scale. The more knowledge that is captured during projects, the more work can be leveraged in other use cases. Of course, this requires a certain rethinking on the part of our customers. Runbooks are easier to understand as the complexity is lower. Conversely, the AI approach is always challenging at first, like any new technologies. We support our customers with training to understand this approach better, so that they can write their own AI building blocks as soon as possible to take advantage of this approach.
There is a lot of noise around automation and even more so around AI.
How do you see Arago fitting into those discussions?
Basically, every provider of AI solutions appears to have a different definition of AI. For example, if you speak of self-learning AIs, these are systems that learn from successes and failures. Of course, nobody in an enterprise environment wants an AI to have several systems broken first, so the fact that it learns from failure is rather a bad idea. HIRO's approach, on the other hand, is to create knowledge building blocks that are applied in specific situations. These WHEN and WHERE operations of an AI must be open enough to be reused, but also restricted enough not to be executed at the wrong time. With these definitions one provides a basic framework for the application of the knowledge and thus prevents unexpected execution.
Arago has found an approach to AI that can also be used directly with customers. One issue that has to become better over time is that customers will have to understand the difference between Machine Learning (ML) and Machine Reasoning (MR). Customers speak about ML all the time while MR is widely unknown. Nevertheless, MR is a much stronger approach than ML alone, but we have to create more market awareness to outline both the differences as well as the benefits.
What are the lessons learned from the early deployments, for better for worse?
HIRO needs knowledge to work. This information is stored in 3 ways: Factual Knowledge (stored in the graph database), Situational Knowledge (Issue) and Actionable Knowledge (Knowledge Items). Creating this knowledge is the first task needed to start automating. This requires a lot of information from the client who wants to automate his environment. If there is no well-kept CMDB to extract the data or no experts available to explain the process, you can’t feed this information to HIRO. Clients need to be aware of this. To automate you don’t just need a good tool, you also have to have the processes and the information available.
You are a veteran in the world of automation. What piece of advice would you pass on to clients?
SYSback is really a veteran in the world of automation. We have had more than 3.000 customer meetings around this topic over the last 10 years resulting in endless automation projects built with several different automation tools, leading into completely different IT architecture and discussions about organizational change. So how do we approach these discussions today?
If you automate “holistically” you are changing your delivery model and thus your company. That means a lot of “change management” since successively everything will change when automation takes over the work of humans. Humans are freed to do more creative work, e.g. root cause analysis or project work instead of incident management or stupid deployment management jobs. Nevertheless, most people do not want to change. Therefore, you have to prepare for a lot of resistance.
So, if I would have to tell newbies only one thing then that they should focus on “creating a decent governance model”, otherwise your automation project is doomed to fail. Therefore, everybody must sit in the same boat and should not be afraid that automation might destroy jobs, otherwise everything will end in resistance of the workforce. Rather, there must be a change process in line with the actual technical project and the C-Level definitely has to govern this with a strong support process.
Thanks for all the valuable insights, Olaf. We look forward to continuing educating the market with you and all our partners. In such a noisy marketplace, practical experiences are ever more valuable!
Article published on › Linked in by Thomas Reuner, Head of Strategy at Arago