155K-Member-Banner.jpg

Companies need to be mindful of the pace of change in technology

Stacy-Irish
Posted: 01/02/2018

Ben Stoneham, co-founder and chief executive officer of robotic process automation technology company autologyx discusses enabling agile business, innovation and the pace of change in technology.

autologyx is an intelligent robotic process automation platform, designed for all sorts of processes the traditional desktop RPA doesn’t handle well.

Can you go into detail about what those processes are?

We set out about five years ago to solve a particular problem in operational compliance that was characterized by really complicated processes, very often involving multi-variant decision making, data persistence where the context and the content was a really important process.  And those were the sorts of processes that had traditionally challenged automation.

There were automation solutions like BPM out there, it tended to have very big-ticket values, big implementation times associated with them, and in the world of compliance where we originally were addressing it, those sorts of industries weren’t very well suited because it was all very fast moving and by the time you’d implemented that type of solution, very often everything’s moved on, and so they never really hit the mark.

And we came up with a novel way of building and implementing processes so the engine could run them in a very interactive responsive way, we solved that problem brilliantly and then when we took a look at the wider world. We found that actually things that characterised operational compliance were just as applicable in any number of other business situations and we found ourselves being used in all kinds of areas outside of that original use case.

And here we are today.  We cover all kinds of different vertical sectors, law, recruitment process outsourcing, legal process outsourcing, retail, a bit of financial services, but all the processes are very much characterized around this same idea of complexity, lots and lots of decision paths, reactive decision making.  It’s exactly the sorts of things that people have always really needed people to do because they’ve been challenging to automate previously.

Has there been a trend you’ve seen in the challenges people are facing?

What we are seeing is that, the groundwork has been done by people like Blue Prism who, I think it would be fair to say, were pretty instrumental in coining the RPA and owning the RPA term.

You know, we’re seeing clients who have implemented those sorts of technology, very often in back office applications, the so-called long tail application, and have found that technology works really, really well for them but are now looking and saying well, actually I want to now build things for things which don’t already have an existing desktop applications and how do I do that because the desktop solutions are absolutely fantastic if what you want to do is just simply take a worker out of a process and replace it with something that replicates those steps.

But what about if you don’t have applications already in place that you want to automate around.  You know, if you want to build something from scratch, that’s not something that the existing technology fully addresses, which is really where we step in, hence our focus on innovation.

Is automation something that should be headed up by IT or the business?

That’s an interesting question.  I think the ultimate answer is it depends on the context, doesn’t it?  I mean, our focus is around empowering business and operational people, so what we aim to do with our platform is actually give those people the tools, give them assistance easy enough to use, so that they can think of a problem that they need to address, or an opportunity that they want to go after, you know, build it themselves in our technology using drag and drop, connect it to all the other things that they’ve got that that would include legacy systems as well some more modern things using APIs and that sort of thing, so that, they can just then deploy that into the business and solve the problem there and then, rather than having to go through the conventional procedure and procure, build, deploy, change requests and everything that goes with that.

I think all too often IT gets in the way of agility within a business and that’s not really the fault of IT.  I mean, IT is fantastic at doing the infrastructure type things, like they have traditionally done, and I’m not really sure that they are always the people to go to for innovation because they’re not necessarily the people who’ve actually got the problem in the first place.  The operations people are, so I think in other cases there’s a great case for IT being very involved.  So, clearly you need to have the support of IT in a lot of instances.  You have people in the business who can facilitate the sort of things you need to be able to do, to assist with those sorts of things.

In a perfect world where everything goes right, what’s the one innovation you’d want to happen over the next 12 months?

One of the things that we say to all of our customers from the outset is that the pace of change, the way in which technology is becoming convergent.  I wouldn’t say the rise of machine learning and AI, but actually the reality is there’s not that much new under the sun in that field.

I think there’s been an awful lot of snake oil being sold when you consider that, for example, the first neural network design was actually something like 1948 or 1950, and really all that’s changed since then is we’ve got a whole lot more processing power and a whole lot more data.  It’s just become easier to do.

What we say to our customers is if you look forward to the next ten years and, as I say, I firmly believe that automation is going to be the defining thing of the next decade, it’s really hard to predict where we’re going to be in two years or three years or four years, and what we aim to do. We describe our technology as a fabric automation platform, where we’re not trying to second guess what the winning technology will be.  What we’re enabling people to do through the use of our platform is to be really agile around trying things out, reiterating through different approaches to solving problems, and addressing opportunities.

If I say anything about what I think in the next 12 months, I think you’re going to see some really good practical user cases, actually some stuff that’s been around for quite a long time that is going to mature within some quite straightforward content classification.  It’s technology that’s actually been around for quite a long time.  The challenging thing for businesses has always been, how do I use it in the real-world context?

The math underpinning an awful lot of it has been around for a very long time.  As I say, neural networks as a principle were worked out in the late 1940s, early 1950s.  To put it into context, I have been working with the UCL Machine Learning department recently, and somebody there told me that when he designed the very first actual neural network to do some image classification in the 1980s, it had 15 nodes, which I think if you compare to today, you’d be into hundreds of thousands or millions of those or whatever!  It took him three months to train that neural network using the available processes he had.

Today, the trend networks have got, millions of codes in the same amount of time so, I think the technology has been there for a very long time.  The actual number of businesses that are out there that are doing genuine innovation in and around deep learning are small.  Unless you’re doing it on DeepMind and a handful of other places, I think most people are building on research that actually has been around for a long time.

The question is really not whether or not you know it but is it actually doing anything useful?  I mean, to take the content classification example, you know, regression, probably nominal regression has been around forever.  It’s nothing new.  As I say, the issue has been how do I as a business person get that into a context and use it and train the model in a practical way in my business?  And I think what technologies like ours, and I’m sure others here, is for the first time make that easy to do, so you start to get something really valuable, some business value from it.

If you would like more insight into RPA and where it can be deployed in your organzation download a copy of the white paper Robotic Process Automation: A New Era of Agent Engagement.

Stacy-Irish
Posted: 01/02/2018