Outsource recently got together with Jonathan Crane, Chief Commercial Officer at IPsoft, for a wide-ranging discussion about automation, cognitive capabilities, his organisation’s groundbreaking Amelia product, and the impact of all this upon today’s outsourcing space and the broader business environment…
Outsource: Jonathan, thanks for joining us. Let’s kick things off with a look at your own background and that of IPsoft: how did you come to work with the company, and what’s been your joint automation journey thus far?
Jonathan Crane: My background’s in communications computing, everything from running a software company in the ERP space to, most recently (and most germane to this conversation) being chairman of the board and President at Savvis, now owned by CenturyLink. Savvis’ claim to fame was that it offered not merely colocation – physical data centre space for customers – but it was also a managed services provider. And that managed services capability necessitated you placing your equipment and your technologies within the data centre owned by Savvis. The personnel associated with the data centre were then used to manage those technical environments, and the economics were made by spreading those people out against multiple customers. The customer then got less costly labour support for managing his infrastructure.
At Savvis I came across a company called IPsoft, and met with the CEO and founder Chetan Dube, and talked to him about the future of managed services: don’t be relegated to managing infrastructure within a data centre, but be able to look across all the environments of someone’s information services, and manage it not only in a data centre but across far-flung territories, and in other people’s data centres, and on other people’s networks, and be able to manage a wide breadth and depth of technologies.
Once they saw the breadth of that opportunity I joined here, and found a couple of things. One, our base business was what I’ll call remote infrastructure management: we provided all these services for everything from networked devices and networks themselves through storage, servers, databases, middleware, custom applications – all the things you’d expect – and we did it in a very unique way. Unique in the sense that not only did we have the expert people but we had developed a technology to basically assess an incident, an event, in these kind of infrastructures that we were managing, and if we had seen it before, and there was a pattern of this type of incident, we could quickly write an automation: drag’n’drop on a decision tree, very simple architecture that we’d pioneered. And we were able to take these kinds of incidents and remediate them more rapidly than anyone, and also by that experience we would build into our knowledge library a virtual engineer that could be used should that incident arise not only in that customer but generically in any customer. We have built some 15,000 of those automations, for anything and everything you could imagine in an infrastructure environment.
That then gave us the idea that if we could do this for ourselves, why don’t we enable service providers and systems integrators with this technology as well? So three years ago we focussed heavily on this enablement, delivering the technology that we use to provide this kind of service. We built this relationship with NTT, for example, in Japan, and Accenture, on the systems integrator side. We have a very broad swathe of service provider partners who are now standardising on our automation technology: we call it IPcenter.
We also have gone up the stack in terms of large sophisticated customers, providing this capability behind their firewalls and working with them to manage the services that reside on those platforms. And then, most recently, last December we announced something called Amelia, which is our answer to the cognitive technologies that have been talked about for years such as IBM’s Watson, Google’s driverless car, and even Apple’s Siri. There have been a number of cognitive efforts that are now looking at how to use a computer in a more human fashion. Our focus with Amelia has been how to do the same kind of assistance work we do in managing IT in business processes as well. We’ve designed Amelia to be able to do the same things that humans do that are repetitive, mundane, tactical: everything from working at a service desk to working within business process supporting a whole range of services using this cognitive technology, with the knowledge that she has of a particular work process or particular issues within a work environment. She’s schooled in being able to respond and solve that kind of issues.
So that’s the progression of automation within IPsoft. We’ve gone up the stack from the automation of IT that automates the interface between machine to machine to now looking at cognitive solutions that automate the human to machine interface. We’ve seen the benefit of autonomics in IT management in terms of lowering cost, increasing speed of service as well as quality of service and are now expanding that experience to improve customer experience in a much broader sense across business processes. That’s our story.
We’re a privately held company, financed by our founder and owner via private family trusts has, and we’ve been growing our business solidly for 16 years. We believe very strongly that we have a right to this new cognitive space, based on our experience and our success with autonomics in the IT environment using our virtual engineers. We call Amelia a “virtual employee”, and we’ve come now to look at those aspects of knowledge work that really would lend themselves to a cognitive solution.
O: OK, let’s take a step back and look at some of the earlier work you alluded to. How much of your work with and for clients is still composed of that more basic automation?
JC: Yes, we clearly are in the early stages of Amelia’s launch – we’re in pre-production environments within six clients. Revenues are coming from project management, project implementation; Amelia’s still very much early-stage.
One thing I would say, however – and I really do appreciate the difference between cognitive and what we’ll call “first phase” automation – is that I look at the automation we provide as breakthrough automation. And the reason I say that is because most people are still doing automation on the primary platforms that come from HP, BMC and CA: those companies pioneered an ability to do what’s known as scripted automation. And scripted automation requires knowledge transfer from an operational person who’s looking at a particular incident, has to categorise it, describe it, has to get it over to a development resource who then goes off and works a development script, and then takes that development script and tests it against his/her operational environment, and then implements it into the environment.
This can take a long time; it can be costly; you have to fight your way out of the priorities of the development team. It’s a very different process. With IPcenter, we’re not talking about the development team making the automation. We’re talking about the actual people who view what’s going on in the operating environment creating an automation using a few drag’n’drops, validating with the specialist team that this automation needs no further technical modifications to run reliably, and then get customer approval to let it run in the environment. Sometimes this is a ten-minute or less exercise. We’re not going outside the environment of what we’re viewing, and obviously we’re working closely with the customer.
This is very different from stepping aside from operational performance and metrics and management, and employing an expensive development team to script solutions. It’s a very different approach – and we continue to enhance it every day. We’re going to enhance our IT management functions using cognitive in future and integrating that additional set of capabilities into the environment.
To continue reading this article, click here…