Digital operations through AI-driven software robots
16 July 2018
Adam Devine, VP and Head of Marketing for WorkFusion
Enterprise companies in every industry struggle with the same fundamental problem: how to unchain revenue growth from cost growth. Adding headcount to do more work and executing custom IT projects to automate some of the work adds scale, but costs grow in lockstep with any increase in revenue. Born-digital businesses such as Amazon and Google have a significant operational advantage through the native use of digital technologies such as artificial intelligence (AI), which drive exponential growth while keeping costs in check. But how can born-analogue companies achieve these margins and improve customer service?
Most businesses have made great strides in digital sales and marketing, but operations have fallen behind. McKinsey estimates that over 60 per cent of data collection and processing can be automated with technology available today, and PwC estimates that AI will drive $15billion in value by 2030. Software bots that learn are the present and future of business growth and will help born-digital companies create digital operations.
Until recently, AI was considered too complex and expensive for enterprise operations, but that has changed in the past five years with the advent of AI-driven robotic process automation (RPA), which seamlessly integrates AI into business processes. Common functions such as customer onboarding in banking, claims handling in insurance and F&A functions such as accounts payable and order-to-cash are filled with repetitive systems work and unstructured data processing that require lower-level human judgment. AI-driven RPA can automate systems work through rules-based bots that perform tasks as humans would on UI of applications or through API calls. Cognitive bots learn patterns from historical data and human actions and automate more complex work. The combination of both automates a wide variety of high-volume, mission-critical work within weeks of deployment.
When companies achieve digital operations, everyone wins. By being freed from repetitive work, employees are able to focus on more meaningful, value-added work and deliver better customer service. Customers win with faster response times and more personalised, accurate service. Company leaders win not only through expanded margins but also with the operational agility to break into new services, products and markets. Unlike electricity, which took 30 years to make its way into factories, AI-driven automation will likely take a third of the time or even less to transform the way companies work.
For information about achieving Digital Operations, visit www.workfusion.com
Welcome to Business Reporter's Intelligent Machines Campaign. I'm Rachel Hicks. For start-ups and data first companies, the deployment of artificial intelligence, machine learning, and big data comes naturally. But what about the old guard businesses who want to increase their efficiency as well as to keep some of their legacy systems? Can they ease their enterprise into a smart process automation journey with minimal risk and disruption? With me in the studio is Adam Devine, VP and head of marketing for WorkFusion.
Adam, good morning. Thank you for joining us.
Good morning, Rachel. It's my pleasure.
Now, we're hearing about AI and cognitive automation in business, but What does it mean?
So let's start with artificial intelligence. Artificial intelligence is the buzz word in business today. AI is basically the capability for software to learn, for computers to learn, usually from people, sometimes from situations, but AI is software that learns, right? Cognitive automation is a part. It's powered by artificial intelligence.
And when you look at a business, the idea is to automate these days. You want to automate as much of the repetitive work as possible. So cognitive automation uses the patterns of work that's been done and how people are engaging with systems and information in real time to find those patterns and train the cognitive automation to automate that manual work.
So what are the fears and concerns of businesses looking to adopt AI and cognitive automation into their business.
Probably the number one concern that leaders have about cognitive automation is that it's difficult to integrate within their business process.
And is it?
No, it isn't. I think maybe five years ago, when the technologies were more nascent, and you needed a lot of data scientists to train machine learning models and choose them and deploy them, it was a more novel, more costly endeavour. But now, thanks to something called process AutoML, Auto Machine Learning, basically, all that work of cleansing the data, the datasets that you need to train the models, choosing the right machine learning model, and combining the two, applying the data, the model-- all of that's been automated. So it's become basically a self-service for business.
So are there other risks or concerns other than the integration of the system that business is worried about?
That it's expensive. It's no longer expensive. Any business can afford some level of AI at this point. That you need lots and lots of data scientists-- not true. Business people can use it. That you need tremendous amounts of data to make it work-- not true. A few hundred repetitions of any kind of process or example of data is sufficient to train machine learning these days.
OK. So we can't worry about the cost too much. We shouldn't be concerned about integration. So let's say we've got all the new cognitive automation in our business. What critical business problems does it solve, and how does it solve them?
So there are two critical business problems that any leader within any business is going to be thinking about. The first is how do I deliver a better customer experience? Businesses have done a great job with digital experience. So the front end of the business-- creating apps, creating beautiful websites, creating live chat-- but the middle and back of a business is still largely in the dark ages. It's still lots and lots of people doing lots and lots of repetitive manual work that's not their fault, but filled with errors and filled with delays.
So the first question is, how do I get my back-end operation to match my front-end digital experience? The second one is that growth is chained to cost. For every dollar that I earn, I've got to spend $0.95 to do it. So what cognitive automation does is it detaches growth and cost. It allows businesses to achieve elastic productivity, elastic scalability by using software bots that scale up and scale down on demand.
So you can, instead of hiring lots and lots of people to staff for your maximum capacity, you automatically scale up bots and people when you have peak demand, and then you can scale down in those troughs of activity. And that, over the long haul, is going to ultimately detach growth and cost.
So does this then mean that you eliminate the need for data scientists?
There's always going to be some need for data scientists. I mean, it's one of the fastest growing fields out there. But what you will not need is data scientists to automate business processes. There's one very specific field that we're talking about here. Granted, every business is made up of business processes. But instead of needing lots of data scientists and engineers to automate these business processes like claims handling and insurance, product categorization in retail and CPG, customer onboarding, KYC, anti-money laundering, and banking. You don't need data scientists to automate those processes anymore.
OK. So let's talk about RPA Express, which is the world's first free RPA product. What does it do?
Let's start with the acronym RPA-- Robotic Process Automation. I mentioned earlier, repetitive processes within a business. RPA is great for work that is truly repetitive, that has no variation. Imagine moving a line of structured data in an Excel sheet to SAP or Oracle. It's basically copy and paste work. So RPA uses rules to move information from one system to another, to operate the UI, the User Interface of applications like SAP, Oracle, Workday, Salesforce-- basically, all of the systems that comprise the bedrock of a typical business. That's what RPA does.
RPA Express is WorkFusion's free product that any person can download, whether they're a business person or an IT person and immediately start using software robots to automate the work that happens on a desktop.
If companies want to up grade beyond that, and they want to upgrade to SPA WorkFusion's AI RPA product, how do they do that, and what does it give them?
The great thing about RPA Express is that it's an on-ramp for digital transformation. It gives one business person a bot that automates their desktop. But typically, once the benefits have been demonstrated by one person, it spreads like wildfire. So RPA Express has a subscription version, a subscription that allows multiple users to use multiple bots on multiple desktops. So it's great for a team. What SPA does is allow businesses to scale the benefit of, not only RPA, but also cognitive automation-- we call it AI-driven RPA-- across an entire business.
So in looking at what AI, cognitive automation can do businesses, can you give us an example of how growth and revenue can be affected if companies incorporate it into their business?
Absolutely. So let's look at a specific example. A lot of our customers are global banks. And a great example of an old guard, established, trusted brand that has really embraced digitization in its truest form would be Standard Bank. They're the biggest bank on the African continent. And the first problem they wanted to solve was customer onboarding. This is the process of allowing new customers to get a line of credit, to open up a checking account, to get a mortgage, and so on.
And if you look at Standard Bank's customer onboarding process, it was 22 days. So the process of taking all of the information from a customer, making sure they were credit worthy, and making sure that they weren't on some sort of a sanctions list, that was a 22-day process. And with WorkFusion, they were able to reduce that time from 22 days down to five minutes.
So imagine the implications for cost. Imagine the implications for revenue. If we're going to double-click on revenue, within that five-minute period when a customer service rep is working with a person to onboard them, by making that process so much tighter and so much more condensed, they can actually upsell that customer. They can learn about that customer during the process of onboarding them. So that's not only a way of reducing costs. It's also a way of opening up doors to new revenue.
If we look at the metrics, the evidence-- SPAs, ROI-- you've given us a great one there. Can you give us maybe something in a different sector, just so we fully illustrated the impact it can have?
So if we're talking about the ROI on cognitive automation, on AI-driven RPA, there's a couple of different factors. First, when you look at the subscription, our customers pay off the cost of the software within months from the benefit of using the product. If you look at just RPA alone, typically, you're going to see 10%, 20% cost reduction on a typical process. Because of the additional capability built into SPA, built into AI-driven RPA, our customers are able to automate 70%, 80%, 90% of a process.
For example, in banking, the reduction of fines-- most of the biggest banks in the world are using WorkFusion SPA in some capacity. And with AML alone, these big banks are spending $1 billion or more a year on, not just headcount, but also paying fines, right? So you're reducing, not only the cost of labour that goes into that process, but you're also reducing the paying of fines.
In another sector like the data world-- Thomson Reuters for example is a good customer of WorkFusion-- there's also another data vendor that had been spending $25 million a year on product categorization.
So one product might have lots of different attributes and lots of different classifications in different markets. And they were spending $25 million a year to harmonise all of the data from different markets about specific products. And with WorkFusion SPA, they were able to reduce that $25 million a year cost down to about $4 million. So it radically reduced the cost it took to do something very fundamental to their business.
OK. So let's say C-level decision-makers are watching this. What considerations should they have or should they take away from today if they're considering implementing and adopting AI-driven RPA? So I would say the first consideration is, don't shoot for the moon right out of the gate. I think, if you look at the headlines over the last three or years, a lot of big companies have spent $50 million, $100 million on these POCs or pilots using cloud-based AI providers.
They've often failed because a lot of people go into it. It's essentially a two-year experiment. You're spending millions of dollars on consultants, and you're trying to do something radically new with a new capability right out of the gate. I would say to any C-suite leader, look first at the guts of your business. Look at the middle and back office. Do the boring stuff first. Because that's where you're going to build your stockpile of cash to be able to invest in those moonshots, those big grand experiments. I would say also, don't integrate all of the different capabilities.
Don't go out in the market with a big shopping cart. Look for one platform that has all of the automation capabilities that you need-- being able to ingest data, being able to automate rules-based tasks, being able to automate unstructured work, common everyday work. And I would also say-- and these aren't my words. I'm stealing them shamelessly-- think big and start, not small, but start at a modest point. Look at 5 to 10 processes across the business in the middle and back office.
But also don't start too small. Don't start with just one process and spend months and months re-engineering the process before automating it. Because a lot of the process of automation inherently improves the process. Think big, start medium-size.
Adam Devine from WorkFusion. Thank you very much for joining us today.
Rachel, it was a pleasure.