Electronic Trading

This autumn, our COO James Wooster was a special guest at the podcast hosted by Robin Amlot, Managing Director at IBS Intelligence. We’re happy to share a transcript of the interview in which they discuss how ‘conversational banking’ and ‘composable enterprise’ come together to shape the future of a more efficient financial desktop. Read the conversation below or listen to the podcast episode 105: How electronic trading is evolving thanks to ‘conversational banking’ here.

James Wooster:

Conversational banking normally has connotations of verbal interactions, a conversation in the traditional human sense. In recent years, the idea of augmenting the human experience with a sort of semi-intelligent agent was added to the kinds of business conversation that you can engage in.

We are now deliberately stretching and potentially misusing that word (conversation) to talk about new artificially intelligent kinds of applications. These look at the behavior of traders or portfolio managers, indeed, users of any kind. And also look at what they are doing with other applications, in order to be able to infer some kind of context or intent that they have in mind.

This way that the application can start to surface information that’s relevant just ahead of when the end user actually needs it. So, it’s a combination of artificially intelligent chat bots and data bots if you like, to make the end users almost superhuman in their ability.

Robin Amlot:

You’re going to deliver me something I didn’t know I needed just before I need it.

James Wooster:

Potentially. In the scenarios that we’ve implemented with some of our partners, we can quite easily predict what data might be required. But what we’re doing is also providing some adjacent data to round out the view and enable the human worker to make the best and most accurate, the most insightful decision. So, yes, it’s a combination of simple prediction and then some very smart prediction going on at the same time.

Robin Amlot:

It’s not the case that you’re actually replacing a human in the actionable process. Instead you’re offering that person more data, more information, to inform them to make a better decision.

James Wooster:

Spot on. What we do as an organization is focus on processes that we describe as being exploratory. In a sense if you can automate the end user and it’s safe to do so from an IT perspective but also reasonable to do so from an interaction perspective, then the technology has been there for a number of years and you can go ahead and do that.

But in many scenarios, either you need to keep that human touch because that what’s your clients will expect or the transaction is so critical that you need to explore the data and the context in order for you to be comfortable that it is safe to make the transaction. And so, at Glue42, we tend to focus on these exploratory processes. So, it’s never our intention to remove the human being because it simply would not make sense in the context of that user process.

Robin Amlot:

Let’s move from the concept of “conversational banking” to “composable enterprise”, which I have to be honest, I initially read as “compostable enterprise” but that’s probably not the same thing at all, so what do we mean by “composable enterprise”?

James Wooster:

It’s funny you said that because other people have said the same thing. So, a “composable enterprise” is a phrase that I believe Gartner, the global analyst firm, invented probably about a year ago. “Composable enterprise” is not a technology, it’s not a vendor-specific thing, it’s a concept and it’s an approach of how enterprises can become a great deal more flexible that they’ve been historically. And one of the interesting things, and that I don’t think even the mighty Gartner predicted this, is that the arrival of the pandemic has forced organizations to accelerate their thinking around composable enterprises.

So, in short, a composable enterprise is an organization that can build and support new business processes and user journeys, very simply, very quickly from the assets they already have or that they acquire. So, for example, if a large enterprise has one monolithic application, changing it to behave and do things differently is a major IT program on itself.

Now, consider the other side of the spectrum, where every discrete task within one of those monolithic applications is like an application in its own right, but it’s an application that understands how to communicate with other applications. So therefore, you can start to roll your own user processes by having these smaller components at hand that you can compose and then re-compose at will.

Robin Amlot:

Can we stitch these two ideas together? The conversational banking and the composable enterprise to come up with a synthesis of what an institution would look like and how it would behave, particularly, and you mentioned it, in the new circumstances we find ourselves in now with the covid19?

James Wooster:

Yes, they go hand in hand really. So, the interesting thing about what we’re doing with conversational banking it that the vendor products that we’re bringing into the desktop are most easily assimilated when those vendor products are delivered in these smaller components.

Let’s take an example of where an application vendor is of the opinion that their application is so important that all anyone does every day is use their application and no other. The consequence of that design assumption would be that it sits on your desktop and occupies the entire screen real estate. That may have been okay for decades gone by, but we’re in a very, very different marketplace even in the last couple of years. And so, the kinds of things that we’ve been asked to do is to assimilate individual tasks and not huge applications. What we’re doing with this conversational banking is to integrate these advanced bots such that they can sit alongside other applications. This is not because the bot by itself is not any good – it’s just that you need transactional systems wrapped around it.

So therefore, we’ve got to the point now where application vendors are realizing that they don’t own the entire desktop. Now, they’re building components, components that are designed for composition. And that, I think, is a really important inflection point on the desktop concept that we’ve been used to for decades now.

Robin Amlot:

How does this evolve over the next couple of years? When do we see this implemented?

James Wooster:

It’s live in many many organizations right now. The interesting thing, though, is that composability almost by definition, requires standards. So, if application vendor A said “Right, I’ve got 20 components that you can reuse but you have to lock yourself into the way that I’ve defined those” and application B does the same thing but in a different way, then composition becomes very tricky.

Now, Glue42 is an integration vendor, we can resolve those issues, that’s part of what we do. But wouldn’t it be better if every software vendor said “Right, we are going to expose our business capabilities in a standard way where they could be discovered at runtime and invoked without any prior knowledge of these components actually existing”. Wouldn’t that be awesome? And thank goodness, financial services as a vertical industry is actually leading the way here.

There is a consortium called FDC3, which is now part of the Linux Foundation. They are defining a set of interfaces and standards that allow software vendors to expose these application components and have them discovered, have them interrogated and introspected at runtime. In the end it’s all of these things coming together at the same time. The combination of these applications and standards and some horrible market forces that are accelerating everyone’s opinion of how the desktop should evolve and it’s a genuinely exciting time at the moment.

Robin Amlot:

But, as you say, it been accelerated, it will continue to be accelerated, I think, for the next year to 18 months at least, until we reach some kind of equilibrium in terms of how people are doing business and how people are getting used to this “new normal”.

James Wooster:

Depending on the industry you look at, I think the current set of standards and the general market conditions have easily accelerated 18 months ahead of where they would have been otherwise and that really opens a whole lot of new opportunities for buy-side, sell-side globally.

If these components, that you can compose and then recompose, have the ability to record their usage then overtime, perhaps over a course of a day, or over a course of a month, or even a year, if you look at the usage of these components, you can understand some very interesting behaviors from a user behavior perspective but also a business process perspective. You can see if a collection of components is being heavily used, while others are not. And if you’re the IT guy or girl, then you can start to consider how you might rationalize your application landscape.

You might decide to remove the lesser used components. In other situations, if you look at how these components are being used and you spot repeated patterns of behavior like component A, followed by component Z, followed by component M, are always taking place within the same kind of scenarios, you might decide to improve the integration between these components because you know they’re always used in that particular order and ultimately these components will help drive further desktop optimization with simply the way they work and the granularity of these components.

That’s something you would find very difficult to do in a traditional siloed monolithic application landscape. So, not only have we accelerated this move towards the composable enterprise, but it will accelerate its own usage by the intelligence and the analytics you can derive from these components.