Co-Founder and CEO – Roots Automation

New York-based Roots Automation is developing digital co-workers capable of completing the work of eight people, CEO Chaz Perera explains how bots are now providing executives with the time to focus on creating higher value for their companies.

CIARAN RYAN: This is CFO Talks and I am very delighted today to be able to introduce Chaz Perera, who’s Co-Founder and CEO of New York-based Roots Automation. Companies are trying to cut costs and are turning to robotic process automation, RPA, and artificial intelligence to do so. Yet most fail because RPA and AI require added layers of technology and technical expertise that are expensive and complex. As a result, some companies try and fail. Others keep throwing good money after bad, and then there are those that won’t even attempt it. Roots Automation was founded to solve this problem by offering zero integration, self-learning digital co-workers as a service. Now, Chaz also worked with AIG and he serves on the board of Habitat for Humanity, so we’ll come to that in a minute. But first of all, Chaz, welcome to CFO Talks and you’re talking to us from New York and we’re in Johannesburg and Pretoria in South Africa. How are you?

CHAZ PERERA: Thank you for inviting me and I appreciate the opportunity to speak with you guys and I’m doing well. I think for all of us, 2020 has been an interesting year, particularly here in New York, given some of the issues we’ve had with Covid but I think things are starting to get to normal. So thanks for having me and I’m excited to share.

CIARAN RYAN: You’re most welcome. Okay, help us understand the problems that companies face when they’re trying to use robotics and artificial intelligence to cut costs. Why are they failing?

CHAZ PERERA: Sure, most RPA technology is very good at replicating mouse clicks and keystrokes that you and I would very naturally do on a computer. They’re also very good at following a structured if-this-then-else statements, but as you and I know, in the working world most problems don’t fall into that nice bucket of if-this-then-else. So to really build a bot that can do more of the work humans do and do it at a high degree of quality with real speed and accuracy requires a lot of other technologies to be brought to the table. AI is one-piece optical character recognition [indistinct] is another. So the ability to read documents and try to transcribe them, as well as workflow and a few other things. So if you want to build a bot that can truly do the work of a human, you need to bring a lot of technologies to bear. Those technologies, quite honestly, are quite expensive, and then you also need to bring people with all the skills that can take those new technologies you’ve introduced and make them work with the bot and that’s really where the cost is. What we typically find is that companies will spend seven or eight months trying to develop a bot and introduce it into their operating environment with absolutely no guarantee of success. So if you imagine spending that much time building something, you’re likely spending tens of thousands of US dollars, potentially even hundreds of thousands of dollars, just trying to get started with, again, with no guarantee of success.

CIARAN RYAN: All right, you’re using bots to solve these problems, you’ve spoken about that. Now, are these bots customised to every client or are they more or less plug and play?

CHAZ PERERA: So they are as plug and play as we can get them, Ciaran. What recognise is that let’s take a finance process like the accounts payable process of companies, what we find is that when you go from company to company, regardless of industry, those processes are largely the same.

I would say between 80% and 90% similar. There is still that layer, that 10% to 20% that is different from company to company. So when I say that we’re largely plug and play it’s that our bots know how to handle the AP process, they understand contextually the language that’s used, they understand the requirements of that job and how a human would perform that job. But there’s still some small amount of configuration that’s required to adapt to a particular company’s rules, hierarchies and, quite frankly, systems, not every company has the same exact, backend ERP system, for example.


Technology that completes the mundane work


CIARAN RYAN: Can you give us a practical example of this? Remember, we’re talking to CFOs and we’ve had scores of CFOs here talking about this very thing, artificial intelligence and how so many of the routine processes are going to be automated for them. Now, of course, they’re all using different packages, some of them have got enterprise software, other people have got packages such as NetSuite, aimed at the mid-tier market. Give us a practical example of where this technology is going to revolutionise their job.

CHAZ PERERA: I think the fundamental thing we as humans want out of our jobs, and if you’re a CFO at a company you likely want out of your team, is to get them to focus on value-added activities. I’d imagine if I were a CFO at a company, my CEO or my peers at the company are worried about activity-based costs. They’re trying to figure out, okay, what levers can I pull, given the economic environment we’re currently operating in and budgets are tight, getting resources is difficult and it’s important that you maximise the time that your people have to focus on those higher value tasks. The way to do that is to take all of the mundane and more rote work off the plates of your people. If you’re a CFO, you likely have people at your company who focus day in and day out on receivables or focus day in and day out on payables. Those types of processes don’t require much human judgment. Data comes in, whether the data is complete or not, the data comes in, and a bot can take that data, massage it, get it into your systems, if it’s a payables process, complete a three-way matching exercise, ensure that payment gets made, that it flows through some type of approval hierarchy and that it appropriately hits, within your ERP system, that it appropriately hits your ledger. Those types of tasks typically are completed by two or three different people at a company. It doesn’t, again, it doesn’t require much judgment. People, quite frankly, don’t enjoy doing that type of work. They want to do things that stretch their brains. So this is the type of technology that if you introduce into your organisation can get your people focused on the higher value tasks so that you don’t have to hire additional people to do this type of work, but also it’s going to help you to improve the happiness let’s call it, of your workforce as well.

CIARAN RYAN: We’re also joined here by NICOLAAS VAN WYK:, who is the Chief Executive Officer of the South African Institute of Business Accountants. First of all, welcome Nicolaas, what do you make of this, the whole move towards artificial intelligence and the way that it’s changing the way that CFOs are conducting their lives and their businesses?

NICOLAAS VAN WYK:: I think originally everybody was a little bit scared and sceptical about the robots taking over but now that a few years has passed and all the hype is gone, I think we are in that phase now where most companies if they are not considering it, they are already implementing it. So it’s good to speak with Chaz and the question I want to ask is what is a typical onboarding process? Is it quite an onerous process, do you go out and do a case study and analysis? So a typical from start to finish job cycle. How long will it take for a company to start seeing the benefits?

CHAZ PERERA: Nicolaas, it’s a great question. I’ll give you a very generalised answer and then I’ll give you the answer specific to Roots Automation. Typically, the onboarding process for these types of technologies can be a few months, often it’s somewhere between seven and eight months, and to see value it typically takes another seven or eight months. So let’s call it one and a half years, you can pay yourself back on that investment and then start to gain additional value through that. Other value that you’ll see through this technology and not simply the cost savings or cost avoidance, but also the quality of the work product should go up. In the world of accounts payable, for example, typically there’s somewhere between a 3% and 5% error rate because humans typically fat finger things, they get tired when they’re doing rote tasks. With bots you can push the accuracy rates up into the 99 percentile. So significantly improving the overall quality of the product, which will make everybody in that cycle much happier. Now, with Roots Automation because we provide bots that are closer to plug and play, and we provide bots that interact with humans and learn and get smarter and more effective at their jobs, we actually can get bots up and running in as little as three weeks, though the typical lead time is somewhere between three and six weeks on average. Our customers are seeing a return on their investment in as little as five months, and typically one of our bots is doing the work of eight people. So you can get to a 250 to 300% return on your investment in a very short amount of time.


Senior executive intelligence cannot be automated


CIARAN RYAN: Wow, that’s blown my mind a little bit there, but Chaz, just looking about five years into the future, what role do you see senior accounting executives playing if so much of their work can be automated?

CHAZ PERERA: I think senior accountants at companies would prefer to be focused and I think typically are focused on how can we reduce costs as a company, how can we enhance revenue as a company. That requires a lot of deep thinking, that requires a lot of conversation and a lot of analysis, and as good as artificial intelligence may be, in the broad sense it is still not good at math. There’s a reason why AI, and you hear that about companies like Microsoft and OpenAI are making huge investments in things like natural general understanding from an AI perspective and very focused on language. The reason is because that’s an easier problem to solve than it is to teach bots how to do very complex math. Now, it sounds strange because computers fundamentally are doing math constantly, right, that’s how they work. But math and science just generally from an AI perspective is not as far along as languages. So even in five years, I do see the role of senior accountants and people in the world of finance being continued to be very focused on those value-added tasks, analysis, providing ways to improve outcome and new revenue opportunities, and understanding how you leverage those different levers to get there.

CIARAN RYAN: Are there any trends that you observe in terms of who’s adopting this technology and which countries are ahead of the curve? I imagine United States, you’re located there, I would imagine that the US would be top of the game there?

CHAZ PERERA: So certainly the US, I don’t know if we’d be at the top of the game, but certainly the US has adopted this technology, particularly over the last three years, there’s been a significant increase in the adoption of technology. If you think about the J curve, we’re on the accelerated part of it. In Europe, the UK and Germany in particular have been leaders in this space and adopting this technology. Then if you get out to Asia, you see places like Singapore and Malaysia being high adopters of this technology. It’s typically places where there’s been historically a culture of rapid innovation, that’s been places where this has been adopted. That being said, this technology is pervasive, it exists essentially every country in the world in some form. More than anything I would say you just need to have an innovative mindset as a CFO or a CEO or COO. If you’re focused on finding ways to innovate, you can very easily introduce this technology and create real value for yourself.

CIARAN RYAN: Nicolaas, did you have a question there?

NICOLAAS VAN WYK:: Thank Ciaran. What comes to mind is obviously competitiveness, I don’t think companies can go without this technology, you can’t run a manual company when your competition is automating. So I think there are great benefits for countries developing national policies, speaking now obviously from a professional body perspective, on artificial intelligence. I see strong leaders there in Singapore and those countries. In South Africa SAIBA is preparing the ground and communicating with the government to ensure that we get ethical robots. There’s so much power behind this technology and there are certain things that we should not do with them.

Just your thoughts on that, I think this there are two aspects, how do you get an ethical robot and secondly, I seen in your LinkedIn profile you had an article about the inherent human biases that you need to avoid from robots.

CHAZ PERERA: I’ll throw this out there first, it’s a book that I’m currently reading, so I’m in the early days of it but it’s one that was recommended to me by a former CEO of mine. It’s by a gentleman called Nick Bostrom and it’s called Superintelligence, and it is focused specifically on this topic, at some point in time the machine brain will be more capable than the human brain, so how do we make sure that we provide that machine brain with the right set of values. I think the challenge is, and it’s analogous to the challenge that companies have when they think about making sure they have a diverse workforce, I think it’s important that the way AI learns is by interacting with humans. One of the things that’s very core to our product is exactly that, that bots can only get better if they have the opportunity to interact with experts in that particular company, in that particular function. The important thing here is to make sure that you give the bots a variety of people to work with, who have diverse thoughts, diverse backgrounds, diverse perspectives, so that you don’t bias the bots. By the way, we’re still in the early days of automated self-driving cars but we’re already seeing how some of those biases come forward, where in the earlier days of self-driving cars, cars were essentially able to see people of a certain persuasion but not of another, and that’s because of the way data was fed to the bot, and so there were more videos and images of people of one persuasion than another. So the car has just naturally said, okay, this is what I need to look for. So it’s important as you build your company that you think about that diversity, you think about bringing a variety of different people together. Well, you need to do the same thing in the world of AI, you need to bring a variety of different people together to help train and build that bot.


‘I certainly think the days of the number cruncher are gone’


CIARAN RYAN: We’re obviously looking at a new type of finance exec going forward, one skilled in areas such as robotics and artificial intelligence and all things technology. My question to you, Chaz, is are the days of the number cruncher gone, are we looking for a different kind of CFO who’s offering something more than just the ability to prepare the books? What are we looking for in a future finance executive?

CHAZ PERERA: I certainly think the days of the number cruncher are gone. I think finance executives need to be technology forward. Ciaran and Nick, you hit on generating value at the company is much more about strategy, it’s about leveraging, understanding cost of capital leveraging, leveraging your resources, whether they’re human capital or financial capital to best effect. To be able to do that in a world where technology is constantly evolving and your competitors are constantly making those investments, it’s really important that you understand technology, you understand how it can be leveraged, how it is implemented. I would argue that you don’t just leave that to the technologists, if you have a chief information officer, a CTO, as a CFO you shouldn’t just be talking to them about the business case or what it’s going to cost to influence something, you should be understanding that to a greater degree what this technology is going to look like, how is it going to improve the outcome for the customer, how is it going to improve the outcomes internally. If we were to spend a little bit more money, would it substantially change the outcome if we were to spend a little less, but it’s substantially changed the outcomes, so thinking about it in that context is really important. I would also throw in that we find often when we talked to CFOs that there’s an ebb and flow to how they work, as you get to the end of the month, you’re very focused on closing your books, and as you get to the end of a quarter, you’re very focused on things like your board meetings and so on. I think it’s really important to just find a few hours in a month, maybe it’s one or two hours in a month where you just sit and stop for a second and read about what’s happening in the world of technology, read about what’s happening and not just in the context of your company and its product and its customer but just more broadly about what’s happening in the world of technology because you will find ways to adopt technologies that on the surface don’t feel like they fit within your organisation, don’t fit the mould necessarily, but that can truly create lots of value for you.

CIARAN RYAN: Where does a CFO acquire these skills? If the days of the number cruncher are gone, then it’s a different set of skills that we’re really talking about here. These aren’t really taught in academia. So we’re talking about how you interact as humans with robots and with artificial intelligence. We’re talking about managing teams, we’re talking about strategy, where is one going to get up-skilled for this kind of new role?

CHAZ PERERA: I definitely think that there are avenues to up-skill through academia. MIT, for example, has a fantastic programme around AI that you can do online and it will give you not just a foundational understanding but really a practical application of these technologies as it’s understood today. So I wouldn’t just simply say, Ciaran, that academia is not supporting it. I think there are a few vehicles there. I think the other is take leadership from people who are out there thinking about this, and a South African who comes to mind is Elon Musk, he’s truly on the bleeding edge of thinking about these technologies, not just for Tesla, but his companies more broadly. What’s interesting is if you follow people like him on Twitter, he is giving you insight and input on how these technologies can be leveraged, how they bring biases forward, and often is pointing you to resources, individuals and information that can help to round out this conversation. So I think it’s

following those leaders and learning from them, not just what they’re saying, but who they’re talking about and who they’re speaking with as well.

CIARAN RYAN: Nicolaas, I wanted to direct this to you, maybe just tell us a little bit about this subject of the changing role of the CFO. If the days of the number cruncher are gone, SAIBA, the South African Institute of Business Accountants does have a CFO (SA) designation, what are these new skills that are going to be required going forward?

NICOLAAS VAN WYK:: We’ve seen over a number of years how the role has changed as the ebb and flow of new developments are hitting the shores of finance executives, they have to adapt, and they are very adaptable. When we interview them we see that most of them are self-made, self-taught, they love what they do but they are not at all number crunchers any more. They are involved with strategy, involved with operations, involved with HR and also with IT. So it’s a towering job function that they have and I think they are some of the unsung heroes within companies. There was research done in Canada, it was called moving from CA to CFO, and it just highlighted 34 competencies that the newest CFOs need to adopt to be able to function in this modern world. So that’s why we’re speaking with companies like Roots and with Chaz to expose CFOs to the new technologies and connect them with the service providers that can help them. Our designation is registered with SAQA, the South African Qualifications Authority, at the highest level, at Level 9. We’re just seeing that to be a CFO you need both academia and experience, and it takes quite a number of years. So we’re building a network of CFOs so that we can get this information to them quicker and in the end I think our end goal is if we have smarter CFOs who are well equipped, they will make a huge contribution to the economic growth of the country, and I think that’s why we invested in this designation.

CIARAN RYAN: Chaz, have you got anything to add to that?

CHAZ PERERA: It’s a great point that Nicolaas is making that the CFO often at a company is either the number two or number three person at the company, and your responsibilities are very broad, the best way to not just to be effective in your job but provide that leadership is well beyond what your title suggests you do and really understand what it is to be a human resources officer, what it really means to be an IT executive of a company because it will just make you as the number two or number three leader in the company, just that much better.


‘They are very much thinking and feeling bots’

CIARAN RYAN: Chaz, we’re running down the clock here, but just a couple of quick questions. If you can tell us a little bit about yourself, where you grew up, your career path, we know that you’re in New York, have you spent most of your life in New York?

CHAZ PERERA: No, actually I’ve lived in a few different places around the world, but I would say New York actually might just be half my life, if not a little more at this point. I grew up around the world and, and also been to lots of different places. New York, for me, for a long time has been a microcosm of the world, you meet people from all sorts of places and all sorts of walks of life, and it’s a great place to get to immerse yourself in a variety of cultures, and that was really what was important to me about staying here in New York. From a career perspective, I’ve done a bunch of different things. I started off in the world of IT at a company called AIG, and I was fortunate to work my way through the value chain there. One of the jobs I enjoyed most at AIG was at one point I was the head of shared services for the company, and it meant that I suddenly got to work with people across the entire footprint of AIG, which at the time was roughly 60 or 70 countries. I really got to understand how different, even though the art of insurance is often the same country to country, the pressure points that people had to apply to serve their customer properly was very different country to country. That led me to think long and hard about what bots need to be in the future. So if you were to ever look at a Roots Automation bot, they are not automatons, they are very much thinking and “feeling” bots, they recognise who’s helpful to them, they understand when it’s most useful to ask questions, when it’s not, when they’re likely to get the best types of responses, when they won’t. That’s very much in the ethos of the company, which is to build these bots that have very human qualities that can not just engage with humans and learn, but can anticipate and adjust because each day will be different, each priority will be different.

CIARAN RYAN: What was that word you used, was it feeling bots?

CHAZ PERERA: That’s right, Ciaran, bots that to a degree have feelings or can feel is what I mean to say. Part of the reason that I suppose psychologically the part of the reason that we like going to work is because we like the people we work with, it’s not simply about money, right? There are other higher degrees of satisfaction we get than just getting paid, and part of it is the people you work with. So one of the things that we strive to do at Roots Automation is build these bots that people enjoy working with.

CIARAN RYAN: A question that I ask everybody who come on here, you’ve mentioned some of the books that you’re reading, one was called Superintelligence, are there any other books that you’d recommend for people, and they don’t have to be necessarily technical books or related to the line of work, what, what would you recommend?

CHAZ PERERA: One that I think, you said not related to the line of work, but I’m going to give you one that I think every business leader should read and it’s by a gentleman named Jim Collins, it’s called Good to Great: Why Some Companies Make the Leap… and Others Don’t. It really talks about how as leaders in a company you have to think differently about how you engage with your teams, how you strive for high quality and understanding your place in the larger ecosystem. My favourite thing from there is before you can run, you need to be able to walk, and before you walk, you need to be able to crawl. Building muscle memory around those activities is really what gets you to success. So that’s something that’s always stuck with me when I read that book and it’s certainly one of those books that I would give to team members and say, hey, take a read.

CIARAN RYAN: It would be interesting to see what the bots say about who’s going to win the election, have you asked them that? Who are they rooting for, are they rooting for Trump or Biden?

CHAZ PERERA: [Laughing] I don’t know who our bots are rooting for. What I would say is it’s clear that nobody knows and it’s likely that no one will know for a few days after the election by all indicators. I think that’s unusual for the US, we typically know either by the time we go to bed or early the next morning who the president elect is. So I think it’ll be an interesting exercise culturally for us in the US to have to be a little more patient.

CIARAN RYAN: It seems almost unprecedented, this could be a very contested election and it may end up being decided in part by the courts, from what I’m reading,

CHAZ PERERA: It does feel like the…what was it, was it the year 2000 with Bush and Gore, it does feel a bit like that. A different set of circumstances but similar.

CIARAN RYAN: What a fascinating discussion, such a pleasure to have you on Chaz, and thank you so much for taking the time out. I know it’s very early in the morning in New York, but I guess you’re an early riser there. We really want to stay in touch with you and I’m so curious to find out how this develops and how this changes the role of finance executives and CFOs going forward. So we really appreciate you coming on and please do stay in touch with us.

CHAZ PERERA: Absolutely, Ciaran and Nicolaas thank you again for your time. I appreciate the opportunity to really evangelize, I don’t just say this because I work in this space, but it is a fascinating set of technology that can really unleash people to do some amazing things and so that’s really what this is ultimately about, we’re giving people back the time to go focus on the things that create much higher value for their company than simply doing something like issuing a vendor a payment.

CIARAN RYAN: All right, great stuff. Chaz Perera who is Co-Founder and CEO at Roots Automation based in New York. Also thank you to NICOLAAS VAN WYK: for coming on and talking to us on CFO Talks.

Ciaran is a seasoned journalist and podcast host. He has a back-ground in finance and mining, having pre-viously headed up a gold mining operation in Ghana.In this podcast he interviews various CFOs, get-ting more detail on the role of the CFO and their daily challenges and solutions.


CFO Club

Become part of a international community of finance executives.