Bank of Singapore has deployed bots in its regional business to drive operational efficiencies, improve decision-making, and “enrich employees’ job scope”, according to its global head of operations and technology.
“Today, about 30 processes are already automated using bots or robotic process automation (RPA),” Eddy Tai, global head of operations and technology at Bank of Singapore, told Asian Private Banker.
“Robots are really good at automating tedious data extraction and data-entry tasks, which frees up our staff to focus on higher value-added activities.”
Specifically, the bank’s live bots are responsible for certain operational tasks — including monitoring data quality, analysing structured reports, and reviewing employee leave processes.
According to Tai, Bank of Singapore’s relationship with bots began in late 2017 with the introduction of Sky, a program developed to generate credit reports. Running efficiently for 20 hours a day, Sky freed up approximately two hours a day for middle-office employees to increasingly adopt a ‘case officer’ approach — taking on more front-office tasks on a case-by-case basis.
“As staff in operations move more towards case officer concepts, bots not only speed up their processes but also enrich the staff’s job scope,” said Tai. “Bots can also be closer to the front office since they handle a wider scope of tasks that are all related to the customer.”
However, Tai cautioned that RPA should not be used to compensate for flawed processes. Not only could bots amplify weaknesses, he said, but by improving processes, bots could prove altogether unnecessary.
“It is important to note that RPA does not fix your processes nor systems,” Tai said, adding that by re-engineering systems, “you may find that you may not need any bots in the process anymore”.
Further, in an article Tai penned for the bank last year — ‘Mapping out a digital future for private wealth’ — he noted that unstructured formats still present a “stumbling block for automation”, leading to the bank exploring natural language processing (NLP) solutions for a number of functions.
“We utilise NLP, usually in partnership with fintechs. We need it to screen documents for KYC,” he said. “We are also using NLP through chatbots, which we have actually been using internally between the front office and operations desk.”
Looking forward, the bank has identified around 56 more processes it intends to automate by 2020, according to Tai, who added that the number of automatable processes “is likely to be finite”.
Across the region, private banks have focused more of their digital resources on improving employee productivity and operational efficiencies. For example, Credit Suisse Private Banking has used live bots since 2017 to handle repetitive and labour-intensive tasks, while UBS has 700 bots globally and 100 in Asia assisting relationship managers in AML, onboarding, and advice distribution processes.