Focus on managing Intelligent Automation in order to scale. Learning from arago’s partner ChoiceWorx.

The immensely sad circumstances of the Covid-19 outbreak will invariably lead to a paradigm shift in the level of discussions around the Future of Work. While efficiency gains and task automation are welcome strategic levers, often around RPA and more recently even around “Hyperautomation”, the focus is likely to shift more toward end-to-end automation and how Digital Labor can help to mitigate the disruption of workforces. These disruptions include employees not being able to get to work, companies using furlough schemes to protect their balance sheets or service providers not being fully set up for remote and home office work. Fast forward to the new normal, what companies will need is an elastic workforce that can adapt to vastly disruptive changes, and, crucially, adapt with little or no prior notice.

When I was sitting down with Sam Gross, founder of our partner, ChoiceWorx, a few weeks back, this was not the topic I had in mind. I wanted to tap into Sam’s brain predominantly for two reasons. First, Sam’s vast experience as a former CTO for leading service providers such as Siemens, Unisys, CSC, and CompuCom, for whom he has delivered complex technology programs at a global scale. Second, possibly because of this background, Sam has a wonderful knack for explaining the most complex issues in a very accessible (and often immensely entertaining) way.

From the biased perspective of somebody who is pondering about strategic marketing issues at arago, ChoiceWorx is a compelling example of a partner who is building their own solutions on top of our HIRO platform. As a firm believer in ecosystems, arago has always proactively encouraged clients to look holistically at the partnership and accelerate the journey together all the way to new business models. At ChoiceWorx Sam has built a portfolio of solutions that all have one thing in common: How do you manage Intelligent Automation in order to ultimately scale the deployments? What I hope readers will take away from our conversation is a more realistic depiction of the current state of play of automation. This would allow organizations to rethink their approach and make progress with challenges that have taken a completely new dimension with the recent and current sad events.

Sam, what are the client pain points you are trying to address and how can you ascertain that Choiceworx gets heard in such a noisy market?

“I think the main pain point really is that the industry is very focused on achieving productivity through RPA and adjacent types of automation. They have embraced this approach having given up on asking IT to create automation on the back-end. If you look at what integration at the presentation layer means, technically I would argue that every single RPA bot has an average of 12 single points of failure. The RPA host, the operating system, at least two applications, each of which is dependent on a web application and database server, plus the network, plus the RPA operating environment and controlling environment etc. Thus, in total probably 12 points of failure. The question is going to be if an automation fails, which they often do, what fails? And that is what loosely spoken translates into broken “bot syndrome”. Bots fail and nobody knows why. That’s the key client pain point. Automation is implemented in order to create productivity, but in the meantime we are taking people out of the front office, yet we are adding people to the back-office to try to operate a technology for which we had no management layer. We are rolling out technology massively to the business, yet we have no proper management strategy for it.

The next question is how does any new entrant get heard in such a noisy market? In our mind this is less complicated than many believe. Here is my view. I have been CTO for many large IT service providers. And in all those years my responsibility was to create and deliver the delivery platform that as a company we used to optimize our delivery services as well as our labor engagements, while optimizing our cost. In every case, I went out into the market just like everybody else. I bought tools as well as built tools. What I was doing was buying motor parts. And I was buying motor parts because what I was trying to deliver is transportation. Today when I think about the AI market, I know that people really want to buy transportation, they don’t want to buy motor parts. So the important thing is that you will be heard in this market only by delivering a solution. We believe that this solution is simplified AI. Namely, AI that reduces labor dependencies. And when I look across the industry, I am surprised when I look at it. Because I believe I am the only one doing that. I see lots of tools, lots of capabilities, lots of creativity. I don’t see a lot of solutions. We focus on not delivering a generalized piece of software but rather delivering a specific solution in the cloud. That is the ChoiceWorx approach.

Talking of noisy market, the value proposition of your Robotinuum product appears to crystalize the current state of the RPA market. Can you talk us through your thinking for launching this product?

I think that when we take our toolset and we focus on launching a product for a specific problem, which is RPA, what we really think about is two things. First, the problem. If there is no problem, then there is no solution. The second piece is that we focus on who is really the user for this solution? If you think about RPA and even Hyperautomation, what is happening is that the user is shifting from the professional, from the “IT professional” to the business user. So that makes you think about how the UI is different. So, when you think about the UI, I would like to invite people to look at the tools we use as we ask ourselves, are these tools really focused on a professional IT user or a business user? The reality is that they are all focused on the IT user, all of them. Lots and lots of information, tightly packed, drill downs etc. All that is really very confusing. Thus, we must rethink as an industry what we are doing.

The result is that all those things that I just have talked about just don’t happen dependably. RPA bots break, so what happens? The automation is not dependable, which means that the ROI and the IT cost associated will suffer. If you really get into what was the thinking around our product, ChoiceWorx is probably the only product that can help with those challenges. Because we think about these issues differently. I looked (and there are many people in the market who are trying to help) we are not the only ones that are seeing the problem, but we believe we are the only ones that address the problem in a fashion that is reliable. And that is in order to solve this problem, we integrate it entirely with a control pane for RPA. That means the UiPath Orchestrator, the AutomationAnywhere Control Room, if that is how we integrate it, we would see only what those mechanisms see. In fact, because they don’t go deeper, they don’t see the cause. They see the result. What we do at ChoiceWorx is we dig in from the bottom up, we go all the way down to the silicon layer and then work our way up to the operating system layer and observe what is occurring through the operating system. So if you really think about it at the end of the day, whether you are an UiPath, an AutomationAnywhere, an Alteryx, an AgilePoint, any of the automation software running, all of it and any of it appears only as an application to the operating system. The operating system sees all of it. Therefore, we focus on a bottom-up approach which admittedly is far more complex and a lot more work. But this comprehensive approach is the only way to understand what is happening to the automations that are running on that host. And that is the thing we do that nobody in the industry has done as yet.

On your journey with ChoiceWorx what are the key lessons learned, for better for worse?

I am a technologist, I love technology. But the key issue is not about technology. I can’t argue that what I have built today is better than what I have built at any iteration of my career. The difference is, where I was building technology before that challenged the quality, comprehensiveness, usability of other technologies in the market, I am not building that today anymore. What we have built today is technology that doesn’t compete with other technology. But rather we have built technology that eliminates human labor. We are really focused on eliminating human labor. Not augmenting human labor. I challenge people to look at the technology in the marketplace, all the tools they can buy, the chatbot tools, the new AI tools, and ask yourself the question, does this tool augment my existing labor, make them more efficient, or does it eliminate my labor? Almost everything in the market today is built to augment existing labor. This is about the transition from onshore labor which transitions to offshore labor, which was basically labor arbitrage, to now no-shore labor. No labor at all. With that I would say the big lesson is that if you compete in the software realm you are trying to answer the question, why is an AI product better than a monitoring product like SolarWinds. And I have been asked that question many times. It is not about the technology, but it is about eliminating labor.

The second lesson was, that when think about our moniker at ChoiceWorx, simplifying Enterprise AI, what the heck does it mean? That means eliminating having to write the big check. Sure, we are a software company, we all want this big check. But organizations don’t understand AI sufficiently to write that big check. They think it is a science project. We have built not only a technology model that eliminates labor but have built a commercial model that eliminates the big check. I would say that those are the core lessons learned. If we dig down a little lower, I think it is a question of science versus applied science, which means we are a fit for purpose platform. We only do certain things. We are not an open RPA platform that you can use to automate anything. We either fix RPA bots or we fix end-user computing. We are very focused, for each of our products. We start with a problem and then we work our way out. That means we understand the data domain that we need to resolve, we understand the operational domain, and we understand the constraints of that domain. And we operate with that. That translates to us offering SaaS, a simple subscription fee, and being fit for purpose for a particular market. We use some of the best tools available in the world to build our platform on, but we give those tools purpose. And that translates giving our customers outcomes. With that in mind the key lesson is, customers want to buy transportation not motor parts.


Tom Reuner

Tom Reuner

Head of Strategy