Former VP at Accenture and EPAM Systems
Daniel has 20 years of experience as a technology consultant working on large enterprise digital transformations. He has spent the last 7 years helping enterprise clients choose and utilise leading RPA vendors such as Blue Prism, UiPath, and Automation Anywhere for robotic-processing automation and AI solutions. Daniel previously worked at EPAM Systems as Global Head of Business Consulting and has spent a decade working at Accenture and IBM leading digital transformation projects. Daniel is currently the Chief Innovation Officer at HealthRede, a technology consultancy firm specializing in RPA and AI solutions. Read moreView Profile Page
Daniel, can you share a short introduction about your background and specific role within RPA?
I’m a digital transformation executive management consultant. I have 20 years of experience; I’ve worked for big firms, like Accenture, IBM, EPAM Systems. For the past four and a half years, I’ve been at HealthRede which is a smaller advisory services firm, focused on developing commercial and enterprise artificial intelligence capabilities.
How do RPA solutions fit in, within the digital transformation landscape?
Machine learning, robotic process automation, computer vision, natural language processing are all the major solution components of artificial intelligence capabilities. RPA is vital, because it has both front-end and back-end components to it; it’s almost like a middleware service, in some respects. They are all important but, as far as customer and marketing adoption of various solutions, RPA is at the forefront, in terms of what I see for commercial adoption of AI technologies.
What’s the median cost saving or efficiency gains for a customer that adopts an RPA solution, roughly?
It varies, but the beauty of RPA is that, of all the AI technologies and capabilities that are coming to market, it’s probably the easiest, in terms of quantifying a business case or a value case. Most people, when they do a vendor assessment and implementation process, because you are either augmenting or replacing human labor, they calculate what sort of returns they are looking for. I’ve seen numbers anywhere from 10% to 50% in acceleration of velocity, as well as cost take out. If you’re talking about a claims processing unit of a healthcare company and they have in excess of 100 claims processors, if you can eliminate 10% or 20% of those resources and replace it with RPA solutions, the business case is pretty clear and demonstrable.
How do you compare the major players in the market?
It’s a pretty narrow market. If you look at other areas of AI, it’s a crowded place. Right now, in the RPA space, there are a big three or four. Those are UiPath, Blue Prism, Automation Anywhere and then Pega is the fourth outlier. Then you’ve got other smaller firms, like Kofax. Right now, in the enterprise landscape, when customers or prospects are evaluating the market space and they are interested in an RFI or RFP type situation, it’s really those big three. I think there will be innovators that, potentially, come to market and join the ranks. Certainly, Pega is climbing up and maybe it’s a big four, in that respect. But it is a small cohort.
How do the big three compete?
They all have niche areas. UiPath is probably the most recognizable brand. They came to market at roughly the same time as Blue Prism, but they have different trajectories. The way I would segment the big four is that UiPath is probably the simplest to deploy and to use. It is very user friendly; anyone who knows how to use Excel can start building bots, in UiPath, with a minimal learning curve.
Automation Anywhere is very different, in the sense that it is more script based, so there is actual coding involved. Whereas both UiPath and Blue Prism have visual UIs, that non-developers can use to create bots. That’s the primary difference between them. I would say that UiPath is probably more user friendly than Blue Prism, even though they are in the same sphere.
Pega is slightly different. They are more focused on the workflow side, at the moment. You can build bots in Pega, as well, but it’s a more slowly-evolved capability. Where UiPath and these other guys started off in the bot development and automation space, Pega was very much more on business process management and workflow. They bolted on the bot capability, as an addendum. But they are catching up and they have a large market presence. For them, it’s pretty easy to upsell and cross-sell RPA solutions, so they are certainly a contender in the marketplace.
Are these solutions all on-premise or are there some that are cloud-native?
They all have different scenarios. You can do on-premise and cloud with all of them. That’s one of the unique aspects of the RPA space, because I feel as if, right now, with a lot of innovative technologies, it’s all cloud first and, sometimes, there is no on-premise component. Within RPA, because of the nature of the technology and the evolution paths, they do have the dual scenarios. I think that is a good thing because everyone is on this relentless march to the cloud and migrating as many apps and services as possible, but there are different levels of maturity, adoption and evolution in the market. Being able to sustain both environments is important, especially in the RPA space, because not everybody, in their market domain, are cloud innovators, so they need to have an on-premise solution, in order to be viable.
How, exactly, does the technology work, in terms of how the software agents are deployed or the desktop or client, to operate the bot?
There are both front-end and back-end components. On the front-end, I would say that it is not very different from an Excel Macro. They have created front-end development scripts that have certain services and APIs. They are, basically, observing your behavior. When you talk about RPA, the process mining component is very important. You can have a user go in and say, these are the 10 routine things that I do every single day; I just want to automate that away. But more often than not, you have a scenario where an organization says, what are my options or what is the value case for deploying RPA? Where are the processes that exist today that are highly repetitive, low skill and can, potentially, be automated? That’s where the process mining is a big component.
Very recently, UiPath actually acquired an exclusively process mining company. Now UiPath has both capabilities in various commercially-evolved forms. They all have some version of process mining, but it’s the degree of sophistication and capability that, I think, is somewhat differentiated.
How does process mining work?
It’s a database process where they are going in and looking at users and steps that they take and surfacing high-volume phenomenon. Basically, it’s pattern development. It’s the same way that, when you do business intelligence reports and analytics, you are surfacing heat maps of common phenomenon or typical phenomenon and you start off in that same manner, where you are identifying common processes. Then you have to do some due diligence and analysis, to see linkages between related processes, to see what is a business process that can, potentially, be automated or augmented with a bot.