It’s hard to escape from the topic of automation lately. At BNY Mellon, clients regularly ask us for insight about what the automation trend means for them, what they should consider, and how they can get started. Whether it be asset owners or asset managers, or companies in industries such as healthcare where transformation pressures are becoming increasingly intense, automation offers true potential for value in the handling of massive amounts of data across a complex ecosystem of business applications. Here are some of the most common questions we’re hearing.
What do we mean by things like robotics process automation (RPA) and machine learning (ML) and how are they different from each other?
RPA or robotics can automate labor-intensive, repetitive activities across multiple functions or systems. We train software to replicate a user's workflow or specific task. Then users will intervene with the software to handle exceptions as they arise.
By contrast, machine learning combines the concept of robotics with artificial intelligence to identify patterns. Over time it learns from those patterns and optimizes task workflow. Supervised learning algorithms make predictions or provide insight to recognize those patterns. Machine learning also has the ability to interpret text. It can also make rule-based decisions and become more efficient without pre-programmed rules.
What are some examples of how either RPA or machine learning are being applied today in financial services?
Banks are targeting areas such as operations or finance that have well-defined, predictable processes and workflows that are often highly manual in nature. Traditionally banks would target functional areas that were not being heavily invested in because these types of solutions can help achieve efficiencies quickly, improve controls and deploy this capability at a low cost.
I've also seen banks use RPA or machine learning in areas such as loan servicing or account processing functions that rely heavily on data entry, or fund management for transfer requests, repair queues, fax processing, etc. This brings efficiencies to the table and also improves controls in their existing environments.
What are the specific efficiency opportunities to improve the outcomes and how does that change workplace culture?
They enable our people to work more effectively and efficiently by streamlining processes and workflows, eliminating manual steps, reducing the risk of an unintended manual error, and improving the risk and control environment.
From a culture perspective, they also enable management to free up staff to focus on clients’ needs or product strategy rather than manual steps or mundane tasks. We can repurpose staff to focus on functions that are more critical. Staff become more fungible and mobile to fill different roles in a different part of the organization.
What are the client benefits?
We've had a number of clients reach out to us to get a better understanding of what we've developed and how we've evolved our programs. Other banks and fund managers have reached out, as well as healthcare industry companies such as hospitals, looking to explore ways to deploy this type of technology to improve their patient experience. We have shared our vision, best practices and lessons learned. Now many clients are trying to launch their own RPA or Machine Learning programs for themselves.
What will the world of RPA and ML look like in 2025?
Integration capabilities and migrating them into your existing infrastructure are becoming more seamless. Many of these tools can "plug and play" with your other applications. Traditionally firms would bring these solutions in as independent tools. Then each of those tools were deployed to support a specific function or product targeting a single problem and trying to solve it.
In the future, we will see more integration within an organization, so that fundamentally new technology creates an entire ecosystem. The focus will be to provide users with multifaceted solutions that solve a variety of shared issues, whether that be better controls, improved efficiencies, or less manual processing. We will have more of a service offering where these tools talk to each other rather than installed independently in your infrastructure. Product areas and various divisions can plug into this and use all these capabilities at will, rather than having to coordinate with different technology groups that own specific tools.
What are we doing now to move us toward that scenario?
One, looking at things more enterprise-wide. We have a tendency to really hone in on the business problem we're trying to solve. We also need to ask who else has the same problem. Should we be looking more broadly to solve it for the organization rather than just my individual function?
Two, we need to be coordinating and partnering with technology to ensure a holistic view of how we want to integrate this service solution, not a black box or silo approach.
Finally, we need more people to identify where these solutions can be used. What's happening now is we have these great solutions and we're bringing them in-house, but we're trying to look for the problems where they can be deployed. Instead, because these solutions are flexible and user driven, the bigger opportunity is to identify the problems and understand the value these tools can add on the whole.
Additional insights provided by Karolyn Ferris, US Segment Head, Asset Servicing, BNY Mellon
Media Contact:
Peter Gau
+1 212 815 2754
peter.gau@bnymellon.com
