How vocal analysis can help deliver more personalized services

Banks and insurers may be able to use intent behind the voice to determine if they can make a sale or not.

The tone of a person’s voice during a call is now being used to determine whether banks and insurers can make a sale or not.

The battle for personalised services has now bled into vocal intent, transaction patterns, even social media behavior, market experts and financial institution executives told attendees of the Singapore leg of the ABF IA Summit 2025 in an afternoon panel.

One company, DataKlout Ai Solutions, is studying the intent in a voice call from an organisations’ customers.

“So while the call is being made, we are able to identify through tonal analytics what exactly the guy’s interest is, if he’s really interested in your service or product,” said Puneet Sinha, founder & CEO, DataKlout Ai Solutions.

This intent monetization, as Sinha called it, can help reduce the number of calls that organizations make. It can also be used in collections, customer service, and fraud detection.

Maybank Singapore, meanwhile, has turned to artificial intelligence (AI) to streamline its chief investment officer (CIO) quarterly market outlook into bite-sized stories that can be more easily read.

“[There’s] accessible insights for the everyday person, not just having to talk to the [relationship managers]; very actionable type of guidance that we've provided. And then for us, we've also seen much higher engagement right across our social media platforms,” said Helena Ooi, Head of Strategy, Sustainability, Digital and Data, Maybank Singapore.

Transaction patterns and social media behaviors can also be used to help banks and insurers onboard new clients, with the use of a gen AI assistant.

Nucleus Software, which builds digital platforms for lending, transaction banking, and digital services business units, uses such data so that organisations can offer“very high personalisation and real-time channels.

In one use case, the company manages inbound and outbound queries of a client using a gen AI assistant, reducing the workload of agents by 50%-60%.

“So with the implementation of this particular solution, using AI, we are able to reduce the workload on the agents by 50 to 60% and they are [able to] focus on another revenue generation stream,” said Pradeep Malik, global head, Digital Services, Nucleus Software.

“This way, the bank is able to reduce the cost by 40%-50%. At the same time, the customer happiness index has gone up,” Malik said.

Meanwhile, Etiqa International Holdings saw a boost in its mobile app onboarding and claims processing by integrating AI in its travel claim assessment.

Today, 60% of the claims are released the same day, said Dennis Liu, CTO, Etiqa International Holdings.

“Before when the customer submitted for claims, it took weeks, or maybe months for customers to get a result. That is a common practice in the insurance market,” Liu said.

It also improved adoption of the Etiqa Plus mobile app. “Just last month we celebrated our Etiqa Plus mobile app having 1 million customers on board,” Liu said.