Exclusive Q&A | Blue Prism CTO David Moss on automation, AI, the...

Exclusive Q&A | Blue Prism CTO David Moss on automation, AI, the future

Robotic process automation pioneer Blue Prism has grown fast under the radar: a British success story – and a North of England success story, too: the company is headquartered near Warrington, halfway between Liverpool and Manchester.

Founded in 2001 by CEO Alastair Bathgate and CTO David Moss, the company floated on AIM in 2016 and now has a market capitalisation north of £1.8 billion ($2.4 billion). Its enterprise partners include Accenture, Deloitte, EY, HPE, IBM, KPMG, and PwC, while customers include Heineken, Coca-Cola, the NHS, Pfizer, HSBC, and American Express.

Last week, the company hosted its annual customer and partner event, Blue Prism World, in London, attracting 1,200 people – double the attendees of the previous year’s event, while the company’s New York jamboree attracted a similar number. Meanwhile, the first half of 2018 has seen Blue Prism more than double its new deals: 450 against 169 in the same period last year.

Internet of Business caught up with co-founder and CTO David Moss at the event for a one-to-one conversation about automation, AI, and future strategy.

Internet of Business: There seems to have been a real buzz about Blue Prism World this year…

David Moss, co-founder and CTO, Blue Prism

David Moss: “It’s always felt like our customers were part of a community that was helping us to build software. Unlike a lot of startups, we didn’t create the software in a garage and then deliver it to the world; we learned what the world wanted first and then worked with them to create what we have.”

So what was the main driver for the creation of Blue Prism?

“There were two things. Firstly, my background is in enterprise software, so in the late 90s, early 2000s, I worked for a typical enterprise company that would build solutions for banks and credit finance companies.

“So I was in the world of waterfall [development], gathering business requirements… but even buying one of those systems takes a year, and then it takes a year to gather all of the requirements, then it takes you two years to build it and deliver it, and before you know it, the world’s changed. You’ve got a piece of software that you’ve paid a lot of money for, but it’s a reflection of the way the world was four years ago.

“The second reason was, when we told business people about our idea of delivering agility to the operation, they were so excited about it, they were so motivated by the fact that we were trying to democratise technology, that we were trying to put technology into the hands of business people. At that time, this was a concept that your average architect would have thrown up all over in terms of the complexities associated with it!

“We saw business leaders yearning to self-serve. We saw technology departments that were creaking under the weight of demand, they were being shrunk down and outsourced, and offshored, and couldn’t do the amount of changes that were requested of them. And we wanted to democratise technology. We thought, ‘Let’s give it to operational people and let them define their own problems and solutions.’”

Back office meets front of house

In many large organisations now, traditional back-end IT maintenance and legacy systems are being brought together under the same umbrella as new product development, front of house. It sounds as though Blue Prism was ahead of that trend?

“Yeah, and what’s interesting about that is that when you’ve got front-of-house, customer-centric systems of engagement, and then in the back office you’ve got your legacy, data-centric systems which you change at your peril, those two systems move at very different speeds.

“Your customer-centric systems are about disrupting the market, about first-mover advantage, launching new products, being the market leader by getting these things out there first. Whereas those legacy applications are the complete antithesis of that – you can’t change them, they’re the rock on which your organisation is built. And so the way I describe Blue Prism, this digital workforce that we provide, is, in a way, as the gears in between those two different architectures.”

You now use the term ‘digital worker’ rather than ‘software robot’ to describe your products. Is there a risk of confusing people who think of a ‘digital worker’ as a human being whose work is enabled by digital technology?

“The point I’m trying to make is broader than nomenclature, if you like. In terms of our vision and where we go with our product, we’ve given organisations the power to create capacity – rather than deliver solutions – and we’ve done that by giving them software robots. But if you think about what a software robot can do, and what image that conjures up, it’s about following procedures, making decisions, doing actions, and so on.

“But where I think we can go as this technology matures, in terms of AI, as intelligent automation technologies start to become more accessible and more consumable, is we’re looking to mature that notion of a software robot into a digital worker.

“But what I mean by a ‘digital worker’ isn’t something that can replace someone in your organisation. It’s about creating a digital surrounding for your employees, if you like. Something that can inform, augment, support, and assist people, to help them access more data than they can see themselves, give them insights that they can’t reach otherwise, help them make decisions and deliver value to customers more quickly. A team of virtual, digital workers to surround them and help them do what they do.”

Augmenting people – or replacing them?

But at what point does something stop being an augmented or enhanced skill and start becoming a human replacement? A lot of organisations are rushing into technologies such as AI, machine learning, and automation in a tactical way, because others are doing it, and – above all – to cut costs, and perhaps to get rid of staff.

“It’s a complex area. First of all, what we’re trying to provide is capacity. We’re trying to give organisations extra capacity in terms of the work that they can do. I think if you look at the landscape of the organisations that we work with, there aren’t people sitting around looking out of the window with nothing to do, there are backlogs, there are shortages.

“I spoke to a financial services institution a couple of years ago and they said, ‘Look, we get 200 change requests a year from our business people, in terms of things they would like to do with technology, and we’ve got the capacity to deliver 10. So there are 190 things that OT are asking for that we cannot do, because as an IT department, we only have a certain amount of budget and resources.’

“So when I talk to my customers who are using Blue Prism effectively to deliver this kind of productivity change, what they’re doing is making their organisations more competitive – not cutting costs.”

AI strategy

You’ve recently announced integration with Google Cloud, Microsoft Azure, and AWS, and with those and other AI platforms, too. What is the company’s own AI strategy?

“One of the perspectives that I have on AI is that it’s a very exciting area, an exciting technology, but how you access it, how you apply it, how you derive value from it, hasn’t been figured out by the vast majority of users.

“I see AI as more of a kind of vehicle to get to a destination. That destination might be intelligent automation, for example.

“Our own AI strategy is firstly about a platform play, so there are organisations like Google, Microsoft, Amazon, that are spending billions of dollars researching this capability. Our strategy is to create the world’s biggest R&D budget by bringing all of those technologies together and surfacing them to business users.

“I want to deliver meaningful AI. I want to deliver that through choice. And I want to allow business people to use something like sentiment analysis, for example, or translation, without understanding json, or Web services, or the nuts and bolts of how it all works.

“So our take on that, if you think about the digital worker we discussed earlier, the difference between a software robot on the one hand, and a digital worker on the other, is that the digital worker has skills and the software robot doesn’t.

“We’ve defined six skills – knowledge and insight; learning; visual perception; collaboration; planning and sequencing; and problem solving. And what we’re doing now is creating this digital exchange that allows us to surface Google’s and IBM’s and Microsoft’s and Amazon’s technologies as plug-and-play skills, or components that business people can embed in their own processes, and use just like downloading an app on your smartphone.

“Those vendors are keen to do that, because it gets them more visibility, but I’m keen to do it because it brings the technology of AI into a more business-like language, into a business-outcome focus.”

Software agent vs human agency

In a number of technology areas, including AI, blockchain, and automation, greater autonomy is becoming prevalent. For example, AI agents that are completely autonomous, self-organising ledgers, digital twins that learn, and so on. Isn’t there a concern with some of these technologies that concepts such as auditability, transparency, accountability, and liability – human agency, in fact – are getting lost?

“Absolutely, I agree. Where AI is helping you by giving you more information, that’s the notion of the digital worker, of surrounding an employee with information. That’s support.

“But there’s a whole different debate when you get onto unsupervised learning, for example, and the kinds of models that are being used in financial services.

“I listened to a very interesting debate recently where some financial services experts were talking about how deep learning and machine learning and those kind of models, from an auditing perspective are going to become very challenging, because there is a fundamental legal and compliance problem of, ‘Who’s responsible for the decision?’ But I think we’re a long way off from that. It’s certainly the right time to be having these conversations about it.”

The UK’s new AI and robotics strategies are positive moves. But is Britain investing nearly enough money in growing the sector, when a single Chinese city, Tianjin, is investing 16 times more money than the total value of the UK’s national Sector Deal for AI?

“As a technologist, I would love to see Britain spending much more on grants and on funding some of these initiatives, but I’m a realist and pragmatist at the same time. We’re not China, and we have a deficit that we’re trying to get on top of at the same time. But I would welcome more investment.

“But I think the UK is in a very strong position. Historically, Britain has been one of the forerunners of computer science. The skills are there, the expertise is there, and the will is certainly there. Even if you look at US companies, such as Amazon, Google, and Microsoft, they are investing in British talent and putting research bases over here.

“As to Blue Prism’s own strategy on that, we’re working very closely with organisations such as Google and Microsoft, but we’re also working with Manchester University, for example, one of the birthplaces of computing.

“We’ve got an academia programme, our own research programme, and we’ve got joint R&D that we’re doing with British academic institutions, because we’re trying to further the respect that we have as a country in terms of being forerunners of this. But the more the government can do to support that kind of initiative, to inject new impetus into it, the better.”

Hitting the hotspots

Your personal and corporate roots are in financial services, and helping companies in that space be smarter and more agile. Where are the biggest growth spots for you now?

“What we’re delivering to our clients is capacity, we’re not delivering a solution or something that’s verticalised.

“There are companies in all sectors that benefit from Blue Prism. For example, Heineken is using Blue Prism to make the organisation more efficient, to drive change, to hook up things like systems of engagement – chatbots – with their systems of record and mainframes. We’ve got organisations like Pfizer, Coca-Cola, American Express uses Blue Prism in the US, HSBC, Bank of Ireland…

“The NHS has got a number of digital workers as well, working in check-in kiosks – airline-style check-ins to help people flow through hospital and reduce queues, and that kind of thing.

“We started out in banking, because that’s what we knew best, and we engineered our software to meet those really high standards from a security, compliance, and governance standpoint, but it’s applicable to any industry.

“Schneider, the trucking company in the US, use us to reduce the number of engine failures they have. It used to be that an error code came through to a driver and he would have no clue what it meant, so he would just keep driving until he got to the next truck stop, by which time his engine’s expired and a million dollars has gone down the tubes.

“What Schneider can now do because of robotics is it can translate the error code back at base, send through a meaningful message to the driver and tell him to stop now, and as a result they have had no engine failures in the last 12 months, versus hundreds in the year before. So you might think it’s trivial, but the business benefit is large.

“So, we’re in many different industries, because we’re effectively selling capacity for people to train in their processes and their industries.”

Internet of Business says

The question on many people’s lips might be, how can a British software company in the vanguard of a new market, with an international footprint, and a market capitalisation of nearly $2 billion, be a relative unknown?

Moss responded, “When you’re a company that’s growing that fast, when you’re standing up in Japan, Australia, Germany, France, the US, and India, and all the areas that we’re opening offices, and you’re building your own infrastructure, then brand awareness follows investment. It doesn’t happen overnight.

“We’re doing a lot of work outside the technology circle to build that brand, outside of the operational and RPA circle, but it’s going to take some time to get that recognition.

“We’re getting some, because of the AIM story and the fact that we’re now a public company. We’re very open about what it is that we’re achieving and what our strategy is. And we have some great investors onboard.”