The future of banking is very likely
to be voice-led. This was a major conclusion at a roundtable about cloud, artificial intelligence (AI) and fintech.
![]() |
| Walton sharing cloud-based trends for AI and ML. |
“We're very excited by voice as a
platform,” he said. “We see that with the advances in technology
underpinned by AI and machine learning (ML) that voice will be increasingly important
and a useful interface.”
Ravishankar, Co-founder and CEO,
Active.Ai, observed that chat apps like WhatsApp and
Facebook Messenger are a part of life for consumers in Singapore.
“The rules of engagement are shifting,” he said. “The rules of
engagement are unstructured. You write... as you speak, as you think.
Customers are getting very used to that.”
Active.Ai works with financial
institutions to provide domain expertise and brand personas for AI
chatbots. The software-as-a-service makes conversational, AI-backed
banking easily available to financial institutions so they can be
more relevant to their audiences, he said. “You start doing and
thinking things in a more natural way,” he said.
![]() |
| Ravishankar says the Millennial customer is a focus for the finance industry. |
“This generation lives on Snapchat, lives on Instagram. They want their bank, their purchasing experience to be very similar,” he said. “Just imagine, you are putting your bank into your WhatsApp. You have to grab the new customers. You can't add 500 more branches.”
Ravishankar added that a scalable,
AI-ready infrastructure that is performance-optimised - such as the AWS platform - is invaluable
for financial institutions as they can test out what they really want
to do. “They (can work) like startups - build very fast, test very
fast and then come back and deploy very fast,” he said. “If you
want to transform fast, this is the way to go.”
While the question of cloud security
comes up often, especially with financial records, Walton shared that
customers have done an about-face in recent years. “(Customers)
have changed to say 'the cloud is more secure than I can possibly be
in my own data centre', that they can't invest in security at the
scale that we can,” he said.
![]() |
| Sharma says AI can protect brand reputations and ensure good customer service. |
Sharma shared that his bank is in the 10th
year of a transformative journey. He noted that customer-facing staff
were finding it increasingly difficult to service customers as the number of banking regulations increase, while banks face
challenges in ensuring a consistent client experience across a large workforce and many
branches in a wide geographical area.
“The experience that you give
customers is very fragmented,” he explained. “Can (each staff) answer
a question equally well and take it to the action and get it done
there and then?”
Sharma shared that Active.Ai's chatbots are currently being tested internally and will give the bank an advantage. “It is important that you are at windows before people start queueing,” he said.
![]() |
| Chong says AI can change perceptions about chatbots. |
Chong observed that chatbots are
thought to be simple, automated ways to answer frequently-asked
questions, but are actually quite complex. Besides looking through a
repository of data to respond with accurate answers to questions and
following preset workflows to complete actions, they can also be
predictive, suggesting additional tasks or products, she said. “AI
can simplify all that and help people get financially educated,”
she said.
FWD's Active.Ai chatbot has not yet
launched, but will be able to improve the customer experience, she
said. As an example, she said some overseas travel medical claims
have as two separate claims, whereas the chatbot would be able to
identify all of the consumer's policies, choose the relevant policy
to be used for filing the claims, and then help the consumer complete
both claims at the same time.
“Consumers don't want complexity,”
she said. “The AI should be as intelligent as a human should be.”
Ravishankar said potential customers
typically have fears about the security, and are unsure how they can
respond to unstructured conversations. While companies with legacy
infrastructure are concerned that their technology is not ready,
digital companies want to know what is possible, he added.
Sharma added that companies may worry
about unsupervised learning, which can be how AI models learn. “They
want to be sure your model is getting trained properly,” he said.
“There must be human intervention to say what you want it to learn
and don't want it to learn.”
The speakers did not feel that AI will
replace the human element at financial institutions. “When the
human is combined with a machine, you get the best output,” said
Sharma.
Ravishankar said, “It is not about
how much cost I can save by removing people, but how I can double my
customer base and do it in two years instead of four years.”




No comments:
Post a Comment