Last month I got horribly confused. Following Infosys Confluence customer event, I made the rash assumption that the then newly announced Mana, was the same or similar to Panaya. If you are familiar with SAP and Oracle landscapes then you may be familiar with Panaya, the Infosys acquired discovery and testing solution that takes time out of upgrade processes through a discovery mechanism aimed at exposing where systems will break in an upgrade scenario. At first blush, Panaya and Mana appeared similar but that’s not the case at all. As I later discovered, Mana has the potential to be much more powerful than I expected. But given my confusion, what makes up Mana?
The easiest technical explanation:
- Infosys Information Platform (IIP)– an open source data analytics platform that enables businesses to operationalize their data assets and uncover new opportunities for rapid innovation and growth
- Infosys Automation Platform (IAP) – a platform that continuously learns routing logic, resolution processes and diagnosis logic to build a knowledge base that grows and adapts to changes in the underlying systems
- Infosys Knowledge Platform (IKP) – a platform to capture, formalize and process knowledge and its representation in a powerful ontology based structure that allows for the reuse of knowledge as underlying systems change
This is interesting in its own right because to me, Mana represents a good example of how what started out as one thing, albeit with multiple uses, in the shape of IIP, has evolved to become the foundation for fresh services that perform entirely new functions. In that sense, Mana is entirely in keeping with the company’s philosophy of ‘purposeful growth.’ But what does Mana do? This graphic gives some clues:
It is better to think of Mana as a way of unlocking value tied up in both machines and people, then applying that to problem solving in the current and future worlds. If that sounds vague then that’s OK because its best application comes from asking questions of systems rather than necessarily having a clearly defined problem that needs solving. What do I mean? In a conversation with Abdul Razack, head of platforms Infosys, he said that Infosys is working with 25 early adopter customers that exhibit characteristics requiring a comprehensive solution of this kind. On use cases, he said:
Currently and primarily around manufacturing because you see, especially around the physical aspects, like the JCIs (Johnson Controls) of the world, they have strategic assets into which we are building automation capabilities. The other area where we see that is in financial services because there are so many applications, there is so much efficiency that is needed, whether it is in wealth management or whether it is portfolio management. All these are customer applications that we manage, and from which we can learn how things are working and then improve the landscape. These are new things upon which we’re working.
Pressed to describe other examples Razack went into detail:
Suppose I am the guy responsible for a KPI for a certain level of product delivery. I want to know if my orders are not going to be fulfilled. Let’s say 20 out of 50 are stuck because of some business or IT process related problems. Let’s say those 20 are held up because of some bug in the system that needs resolution. As a business user, what would I do? I would create an L2 ticket because somebody will go change the configuration and a L3 ticket for some code change to fix that bug. That’s inefficient. We want to collapse that L2, L3 scenario and be able to predict when something will go wrong as well as providing the user with an easy way to traverse systems. By the way, in doing that, the SLA-based L1 situation disappears because you’re now tackling an entirely new class of problem that seeks to discover and eliminate root cause failure. Mana captures what is happening in all parts of the system – hence the knowledge part – and learns about the various ways these complex systems interact – the machine learning piece – and so if you take that one step further and out into the physical world, you can go way beyond simple systems monitoring. You can predict when there will be a failure in a physical asset. That’s the artificial intelligence piece.
The use cases Razack described are surprisingly common. How many times for example have you heard plant engineers talking about repetitive failure? How many times have customers wondered why banks demonstrate a lack of understanding about your account activity, or assets they hold on your behalf? Shouldn’t they know better than you? The answers in each case are ‘all too many times’ yet traditionally, it has been difficult for software vendors to come up with much more than technical systems monitoring. Mana takes this to an entirely new place.
We have seen examples of predictive maintenance in the past but those have tended to be one off or custom applications designed to solve a very specific problem. Infosys is hoping that by developing a broad based platform solution and in partnership with early customers, that it will create an adaptive packaged platform that can be used in a wide variety of cases without requiring significant customizations. Getting away completely from customizations will be difficult in complex environments but having a platform that, out of the box, does many of the routine functions associated with problem step represents a big step forward.
These are early days so we will need to follow up and see how customers are getting on with a solution that is certainly ambitious and audacious. And we like that.
You might also like
Image credit - graphic via Infosys, image © Andrey Burmakin - Fotolia.com
Disclosure - Infosys , Oracle and SAP are premier partners at time of writing.