Ensuring the safety, soundness and reliability of the global financial markets is the central tenet at DTCC. In its pursuit to deliver initiatives designed to reduce costs and risk for the industry, DTCC is exploring the use of disruptive technologies like robotics, machine learning and artificial intelligence.
Gordie Sands, Executive Director and Co-Lead of the Office of Fintech Strategy, sat down with DTCC Connection, to discuss robotics process automation (RPA) and DTCC’s strategy to leverage the technology and unlock its benefits for the financial services industry.
DC: What is Robotics Process Automation?
GS: RPA is a type of software that automates the repetitive tasks that people perform today on their computers. RPA performs tasks that involve actions like opening and editing files and folders, filling out forms on web page, etc. Unlike simpler macros or desktop scripts, RPA can manage fairly complex workflows for process that span multiple tasks and systems.
DC: How is RPA different than artificial intelligence?
GS: RPA software robots (bots) are sometimes considered “dumb” automation, in that RPA bots perform only the specific actions that they are programmed to do following a set of narrowly-defined rules. Artificial intelligence, or AI, including cognitive agents or “smart robots” can, by contrast, evaluate conditions or data points and “choose” among a set of options following a more broadly defined set of algorithms.
DC: How do you create an RPA software robot?
GS: With RPA you usually create a software robot using a visual design tool to lay out a workflow diagram and then configure the steps in the process to perform the work. RPA then creates the application in the background using a higher-end programming language, typically .Net. This enables RPA to create sophisticated workflows that may include a wide range of operations and multiple computer systems.
DC: What benefits can RPA deliver to the financial services industry?
GS: One of RPA’s key benefits is improving functional operations and creating cost efficiencies. RPA often automates repetitive manual processes, allowing organizations to redirect staff to other more strategic and innovate work that requires human judgment and discretion. In other words, it frees up the organizational capacity of an organization.
Say we have someone whose day is largely taken up performing numerous routine tasks. If we can automate some of that work using RPA, we enable that person to be more efficient and create capacity for him or her to learn new skills and take on new challenges.
DC: Can you give examples of where DTCC is pursuing RPA?
GS: One example is onboarding clients of the DTCC Global Trade Repository. Roughly 90% of that process is manual data entry of client information. By automating that part of the process, we can increase efficiency as well as drive higher quality by reducing the risk of human error.
We’ve also automated much of the monthly billing reconciliation process in Finance. Although this work isn’t performed every day, it is effort-intensive and important to both DTCC and to our clients that we get it right. By automating this process using RPA, we are significantly reducing the time and effort it takes to complete, improving the accuracy of the billing reports, simplifying Audit review of the reporting, and facilitating faster and more effective investigation and resolution of discrepancies. Importantly too, this makes the billing process more client-centered, helping over time to promote higher client satisfaction.
DC: How are we addressing cybersecurity concerns with RPA?
GS: We are partnering with DTCC Technology Risk Management on this journey to make sure we are putting forth the best security model. As with our people staff, we set up data access permissions to limit who can see what in which systems, both internally and externally. The security model is based on creating new identities for bots and granting the specific and limited entitlements to control their access to systems and data, just as we do for people.
For example, in developing RPA with GTR Onboarding, we give only carefully vetted and narrowly limited data access to robots, following the same access control process that we do with our people staff. With RPA, however, we are able to create separate, individual bots to perform the range of tasks that a single person might perform. In this sense we can restrict a bot’s access and entitlements more narrowly, strengthening the security model. This is a highly secure framework. Because a bot never needs to think on the fly or make intelligent judgments, it’s easier to restrict and control access.
DC: What is DTCC doing with RPA in 2018?
GS: So far this year we have been finishing up the initial pilot projects and working with the business areas to launch them into their day-to-day operations, and have begun gathering and assessing a range of opportunities to select for our next set of projects.
We have just completed a set of workshops to create the plan for the remainder of the year. At DTCC, we have a very large program around automation, including not just RPA but also other automation technologies like AI. We are looking at how to integrate RPA with our other automation initiatives and other parts of our IT strategy, as well as our long-term vision for services. Our long term goal is to create an ongoing and sustainable capability using RPA.