The UK-based merchant banking business Close Brothers is enhancing its RPA by extracting data from unstructured documents to not only enable end-to-end process transformation but also provide consumers with peace of mind. This is due to the lessons gained from the pandemic and the rapid advancement of AI.
Since embarking on its robotic process automation (RPA) journey over six years ago, Close Brothers has mostly worked on automating structured, rules-based operations. However, the 2020 COVID-19 outbreak brought automation related to unstructured input, such as emails and documents, into the public eye, according to Stephen Durnin, the company’s head of operational efficiency and automation.
According to Durnin, “if customers are having trouble meeting their financial obligations, we provide support, which we call forbearance within the banking sector, in the form of a payment holiday.” We needed to be able to provide help quickly and in large quantities because there were a lot of clients on furlough at the time. We created a scalable automated procedure to evaluate and approve forbearance requests, which we connected with chatbots and self-service webforms in order to provide our customers with efficient service and assist the contact center by diverting volume from our telephone channel.
The pandemic’s effects caused Close Brothers to change its strategy on automation. Its first efforts were aimed mainly at task automation, with the goal of automating tasks and increasing the scalability of operational procedures. However, Durnin claims that after the pandemic struck, the business had to take a more comprehensive strategy, which included utilizing AI-powered tools to assist it get over the limitations of conventional RPA. However, getting there also meant reconsidering its current procedures.
Combining strengths
The corporation combined its automation and operational excellence teams in 2022. To facilitate comprehensive “process transformation,” the Operational Excellence and Automation CoE developed a set of tools and techniques, such as dashboarding, Lean Six Sigma, RPA, low code workflow, and operational excellence frameworks.
“This was really bringing home the point that we want to make sure we get as much throughput within the automation solutions as possible, and that we need to lean these processes before we automate them,” Durnin adds. For instance, when a manual process reaches a bottleneck, it will continue to do so even after it is automated. Creating a flow throughout the process is crucial.
In order to assist with end-to-end process orchestration, it onboarded OutSystems, a low code platform driven by AI. It connects with UiPath, an AI-powered RPA platform, and operates on top of Close Brothers’ historical technologies.
Durnin says, “We’ve developed a vehicle to deliver at scale.” We’re automating ten procedures every two months instead of delivering a few at a time. This is accomplished by developing reusable automated components, demonstrating agility in your delivery strategy, and employing an efficient approach to process discovery and simplification.
Advantages of the ideal alliance
Utilizing UiPath Document Understanding in the company’s wholesale finance department was one of Close Brothers’ first significant initiatives after partnering with the RPA startup UiPath. The company was having a lot of problems before the project. The majority of the papers involved were semi- or unstructured, making them challenging to automate, but manual procedures were limiting the service’s scalability. Because trades had to wait for administrators to process them, payouts had a lengthy lead time. Errors or missing information in customer-provided documentation made this worse, necessitating more effort later on.
According to Durnin, “We work with spreadsheets, email inboxes, and invoices that come in via PDFs, which usually drive a manual process where somebody has to review the document or email, reconcile those documents to make sure they have all the required information, and then go back to the customer if some information is missing.”
Close Brothers may be able to automate the end-to-end payments process and the reconciliation of customer-submitted documents by combining Document Understanding with RPA. Data from unstructured, non-standard documents, such as PDFs, can be extracted using Document Understanding and fed into an automated queue. The procedure might automatically ask the client for any missing information. After that, the ops team receives all of the information and reviews it.
The group began with a proof of concept that had three main goals:
- Examine how well the AI module performs tasks like precisely extracting and balancing data from unstructured papers.
- Establish a procedure for the solution’s monitoring and training.
- Create a recurring set of security and privacy-preserving guidelines.
Durnin’s team decided to put the module into production once it passed the Proof of Concept, which included training it on roughly 1,000 papers and achieving accuracy scores ranging from 95% to 100%.
According to Richard Gregson, commercial director at Close Brothers Wholesale Finance, “the initial procedure had been a very repetitive task for the majority of the team who work in the sales support function — checking vehicle registrations, invoice amounts, and dates, among other items.” “The team, who are very adept at delivering funding and obtaining title, can now concentrate on more of the major risks associated with this process and go over some of the most important distinctions in more complicated transactions thanks to the implementation of Document Understanding.”
Gregson reports that the team has embraced automation with enthusiasm and that a key factor in the success of the project has been the integration of robotic operations with human in-the-loop evaluation.
In addition to its Wholesale Finance business, Close Brothers is expanding the use of Document Understanding. In order to give its contract hire clients a quicker, more effective payout service, the banking company enhances RPA with Document Understanding and offers lending options to independent contract hiring rental businesses. According to Durnin, Document Understanding will also enable it to access additional features in the UiPath Suite, such as GPT integrations and communication mining.