On 23-24 June, British Robotic Process Automation (RPA) pioneer, Blue Prism hosted their global enterprise automation virtual event, Blue Prism World 2020. The idea of this event was to explore how leading financial services organisations are staying ahead of the game through the adoption of Intelligent Automation (IA).
Central to the discussion that took place during this event was the issue of Covid-19, and how will affect the adoption of IA. Many people were asking the question: will Covid-19 act as a watershed moment for the sector? One of the webinars that addressed this question was the Banking and Financial Services Customer Panel.
The panel was hosted by Adrian Ewer, VP Financial Services EMEA at Blue Prism, and the panelists consisted of:
- Jonathan Kidd – Head of Digital Operations & Robotics, Bank of Ireland
- Julian Gray – Head of Automation Service Delivery, HSBC Global Services (UK)
- Steffen Graf – Managing Director Processes & Technology, Commerzbank AG
Intelligent Automation in Banking in 2020
First up for discussion, was how each of their companies were approaching the automation revolution. In addition, they each illustrated how their respective Centres of Excellence (COE) were helping them to do this.
How are they approaching Automation?
Jonathan Kidd explained how the Bank of Ireland’s approach was to tackle automation as if they were an SME or a startup – a “garage environment” approach. This enabled them to get started with this project as quickly as possible, and allowed them to approach it in an innovative way. As things have started to ramp up, and the rest of the banking business has started to rely on the work that they are doing – and they have had to start to take things “a bit more seriously”.
What area did they initially focus their efforts on? They looked at a banking product, a deposit product, that was digital product, but with a manual back-end. As demand of this product increased, so did the need to scale the product. As Jonathan explained, “There was an opportunity there, to use a robot to do a relatively simple activity, but at a high volume,” this allowed them to continue to scale, without dramatically increasing processing or waiting times.
Julian wholeheartedly agreed with Jonathan on this point. This combination, of a manual back-end process, with a resulting long waiting time for their customers, was something that HSBC also wanted to be tackling. Their particular issue was with regards to the process of formal auditor letters. This was something that involved different custom systems through different parts of the business, and the previous (manual) way of managing all of this was not up to scratch.
Centres of Excellence
Steffen explained how their board had asked them to build up a CEO in order to scale-up the work in shared services that they were currently carrying out, “100 people now working on the COE, in different roles […] but we are serving the whole Commerzbank Group.” He stated that he believes that real gains are to be had by utilising a COE approach, because it means that the team(s) working on automation are working closely together with the rest of the business: “We understood the business processes […] we could work close together with our colleagues during the [automation] implementation phase.”
Jonathan, whilst full of praise for the COE model, wasn’t sure, though, that this was necessarily an ‘end state’ way to run things. Indeed he asked: “Now that we have standards, can we take our people, and get them to work at the CEO levels of quality, but whilst being embedded into businesses around the group?” He thinks that it is definitely possible, and that you have the potential for “real value” in people building up expertise in different pockets of the business. For now though, he confirmed that their model still remains “90% COE” and that so far, they forays into the world of the federated model, are like “dipping their toe in the water.”
Therefore, all the panelists were in general agreement as to the benefits that the COE model brings, with regards to shared standards, raising levels of quality across the board, and being able to work on dynamic projects whilst still working closely with the business-focused departments of the company. However, there was also a consensus that their approach will need to change and adapt to the new ways of working in a post Covid-19 world.
Synergies with Automation – When to Use AI?
Next, the conversation turned to the potential synergies between automation, AI, and emerging technologies (such as blockchain & cloud-based technologies), and their particular use-cases. The two main such areas under this topic, were cloud-services & advanced AI usage.
When and where to make use of AI
All panelists were of the opinion that there is no need for AI to be rushed into every part of the business itself. As Julian stated, “Tools like IoT just aren’t on the radar, for people in his position and area of the business. The AI that they do have – it’s in analytics, rather than in the automation area – not likely to change anytime soon.”
Jonathan partially agreed, reasoning that it should only be used where it obviously provides a benefit, in a narrow use-case scenario. For example, using other tech with AI, as part of your overall strategy. He gave three particular use-cases to illustrate his point. Firstly, he spoke about the issue of ‘data-in’, in that “the quality and success of automation is heavily reliant of your input data. AI can add value by providing better quality input data.” This is something that Jonathan felt that the current channels and solutions don’t do well enough at present.
Secondly, he turned to the issue of processing & the decision-making power of AI. This is because “RPA doesn’t make decisions, it doesn’t learn from the data & build up a pattern over time. It executes exactly what you tell it to do.” Therefore, AI can be used to make intelligent decisions instead, based on past performance, in order to manage all this data. This would be particularly useful, Jonathan illustrates, in the bigger use-cases where rule-based systems and decision-making systems both overlap.
Finally, Jonathan explained how he feels that AI can be used in the area of Data Aggregation, where they prepare data in big projects to be fed into a decision-making process. AI can be use to assist when they pass this data through “an analytics data lake.” As the above examples demonstrate, Jonathan clearly sees a lot of potential overlap between AI and the automated processes that they are already carrying out at the Bank of Ireland.
Utilisation of cloud-based services in banking
Additionally, the topic of cloud-services was discussed, and how that technology forms a “synergy” with automation and other areas of their respective businesses. As Julian put it, “cloud is the elephant in the room when it comes to all of this”
And what does this elephant in the room mean for banking as a whole? Whilst Julian thought that it was something that needs to be addressed, and kept an eye on, he doesn’t think the wholesale implementation of cloud-services is necessarily right around the corner. According to him, this is because “There is suspicion about security and bandwidth implications,” as well as “we have a lot of platforms on-prem. Having a cloud service talk to on-prem services, just isn’t a reality.”
The landscape then, will clearly have to change first, in terms of how automation is being run from the ground up. As Julian went onto explain, however, this shift might be happening sooner than we all expected, due to Covid-19. Products such as Government-backed loans, brought in specifically to deal with this crisis, are pushing and challenging the banking industry like never before.
This is due to the exceptionally quick turnaround times that are expected for such products, as Julian explained, “we don’t usually operate to this kind of cadence. Projects take longer than 3 days normally.” As a result he could see HSBC “dipping their toe into the water” when it comes to the cloud, in the coming year.
Steffen & Jonathan agreed on the whole with Julians views, primarily those around the fact that you need the right use cases for this implementation, and they are perhaps not quite there yet – especially if everything is still ‘on prem’. However, they both feel that their respective companies are thinking ‘cloud-first’ these days, and therefore if there is an opportunity to improve a service or offering by utilising the cloud, then companies are more than ready to do so.
Steffen highlighted potential Blockchain use cases, that he found interesting, although these are not really linked with automation. Rather, his stated interest was more in what machine-learning can offer – smart automation. As for Jonathan, he highlighted 3rd party Saas, Amazon Microservices, and GDPR implementation as potential use cases for cloud-services in the near future. Indeed, one of Jonathan’s final conclusions was that, when it comes to cloud-based services in general, we are “seeing frightening growth in that area.”