OCBC Bank has rolled out a generative AI-powered training programme for its 900-strong wealth advisory force in Singapore.
The six-month programme uses simulated client conversations to train wealth advisers across competencies, including investment advisory, client management, product knowledge and wealth planning. Instead of relying on scheduled role-play sessions with supervisors, wealth advisers can practise on demand and repeat scenarios as often as needed.
Traditionally, new hires depend on one-on-one coaching sessions that can last up to two hours and are often limited by supervisors’ availability. This restricts how much practice wealth advisers get and can lead to uneven feedback depending on who is coaching, shares Bob Ng, OCBC head of personal and premier banking Singapore, at a media briefing.
The generative AI tool addresses both problems by playing the role of a client and assessing each session through speech-to-text analysis, evaluating wealth advisers on what they say, how they say it and whether they stay within regulatory boundaries.
After every practice run, wealth advisers receive a detailed report that flags where they fell short, whether in tone, word choice, pacing, or compliance. Supervisors receive the same report daily, allowing them to focus their coaching time on specific gaps rather than re-covering basics from scratch. Branch managers save an estimated 30% of their time as a result.
Within three months, wealth advisers who completed the programme booked twice as many client appointments as peers who had not yet gone through the training, and recorded 50% higher revenue compared with their own performance before the programme, shares Ng.
See also: Asia regulators raise scrutiny on banks amid Mythos AI fears
Building the programme was itself a substantial undertaking. Ng's team spent a year assembling supervisors and top-performing wealth advisers to review real-life client scenarios, identify best practices and design the assessment criteria that now power the generative AI tool.
That groundwork does not stop at launch. Supervisors meet monthly to feed new customer scenarios and updated industry and product information back into the system. "We will always refresh our customer scenarios into the AI to keep ourselves up to date," says Ng, noting that sudden market-moving events (such as a policy shift or an unexpected geopolitical development) can render static training obsolete overnight.
The generative AI-powered training programme will be extended to Malaysia and Hong Kong at a later stage, with content adapted to each market's products, regulations and customer expectations.
Asked whether AI could eventually take over the wealth adviser's role entirely, Ng was unequivocal. "I would rather partner with AI than use AI to replace people. Customers will still want to talk to advisers, rather than a robot. It helps us to be more efficient and effective,” he says.
