Since pilot trials began in February, the chatbot has handled more than 120,000 conversations. About 4,000 (corporate clients, mostly small and medium-sized enterprises) are now using the service each month. Customer satisfaction scores improved 23% during the trial period, according to the bank.
“Our latest iteration of DBS Joy represents a major leap forward as it offers instant and intelligent support to customers, while enabling our SME relationship managers to deliver more focused and value-added engagements,” says Chen Ze Ling, group head of corporate and SME banking at DBS.
DBS has embedded multiple safeguards into DBS Joy. Responses are filtered concurrently through several layers of internal checks based on rule sets and parameters designed to mitigate risk. Following the interaction, experienced customer service agents designated as DBS Joy evaluators assess the conversation quality and flag areas for improvement to refine the underlying knowledge base.
When human assistance is needed, service specialists use an AI copilot that retrieves relevant data and generates recommendations in real time. “This reduces employees’ effort and allows them to focus on work that involves human judgment to address client needs,” says Welson Jamin, DBS’s group head of operations.
See also: AI’s power problem is being overthought, says Microsoft’s VP for energy
DBS plans to expand DBS Joy’s capabilities and introduce the service in Hong Kong and India. The bank has made AI adoption a core pillar of its technology strategy, earning recognition from Global Finance in October as the world’s best AI bank.
