realestatebusiness logo

Subscribe to our newsletter

AI will ‘augment’ rather than ‘automate’ customer service

A leading figure from the machine learning, artificial intelligence and business consulting arena has suggested that machine learning and big data technology will “augment” rather than “automate” customer service roles, such as those in finance.

Speaking during a session on “The Rise of Artificial Intelligence in Financial Services” at the SWIFT International Banking Operations Seminar (Sibos) in Sydney this week, the founder and CEO of data insights and business consultancy GRONADE, Tomer Garzberg, suggested that while big data and artificial intelligence (AI) would provide greater efficiencies in financial services, they would not fully replace human-based customer service. 

The company CEO, who has worked with major banks such as Westpac and NAB, said: “The interesting thing we always look at when we come in to organisations is to understand why things are the way they are. While its OK to band-aid things with a piece of technology, there has to be a fundamental shift in the way we do work and execute work and the work you choose to do fill your eight hours a day with. Typically, that starts with operations.”

Mr Garzberg gave the example of looking at “2,000 customer service agents and what they are doing”, and he found that “a lot of the time, they are answering questions they’ve answered in the past. Why is that happening? They are burning money”.

The CEO suggested that once processes are deconstructed and mapped, businesses, such as banks, can understand where technology or AI could play a role and assist, rather than displace, employees. 


Picking up on a theme postulated by ANZ CEO Shayne Elliott during the Sibos opening plenary on Monday, Mr Garzberg suggested that customer-facing services would not be replaced by digital innovation.

He elaborated: “There is never going to be a truly automated customer service team. It’s going to be about augmentation. And the augmentation needs to come from letting the technology pick and choose what it has to do to service customers better, and then relinquishing control to a human when things get too hard.”

However, Ramneek Gupta, the managing director and co-head of venture investing at Citi Ventures, said that customer services is where machine learning would likely be first applied, as can already be seen with the introduction of intelligent chatbots.

“The first place we’re seeing the application of machine learning is in customer engagement, customer onboarding, customer outreach, customising websites for individual customers,” the MD said. 

“That’s where the industry has the best and the cleanest data.”

The panel went further to suggest that the “blue-collar jobs” of the future will be about “cleaning data” for the introduction of AI capabilities in the future.

The Citi Ventures MD further suggested that harnessing data coming through chatbots and natural language processing (NLP) programs is already taking place and could be of growing importance for those in the financial services sector.

“The importance of having good customer SAP and NPS information is critical to all of us,” Mr Gupta said.

“We recently found ways in which you can leverage NLP to all sorts of incoming engagement with customers, be it live chat or customer service call logs, Twitter data etc. 

“[You can then] find out what the [customer] sentiment is and then correlate that, quickly, with fast information and predict exactly on [date from] 80–95 per cent of your customers versus getting 5 per cent of people to use the [survey] service.

“That’s a very interesting case of applying NLP in an automated fashion across all of your customer base to get that information.”

Ralph Achkar, State Street’s managing director of digital product development and innovation, told delegates during the panel session that while there are systems in place to capture and harness data and support AI, “we’re not doing AI now”.

“What we’re doing is working towards it,” Mr Achkar said.

In conclusion, the founder and CEO of GRONADE suggested that rethinking how data is used and applied to products could be used to create personalised home loan rates. 

Giving the example of his work with a fast-moving consumer goods company, where his company was reportedly able to “optimise the amount of discount that the product needs to have at store level to sell the most amount of volume per year”, Mr Garzberg said: “You can take that same formula and apply it to home loans.

“You [can] look at things other than just dollar value. Volume value is a complete flip side [way of thinking]. So, sometimes… you have to come in and shake stuff up. We call it plastic engineering. You come in and you break things down and rebuild them the way you want them to be.”

[Related: ANZ CEO reveals the ‘biggest challenge’ facing banks]

AI will ‘augment’ rather than ‘automate’ customer service

Latest News

Islamic Bank Australia has been granted a restricted banking licence from APRA. ...

In a move partly focused on guiding its approach towards diversity and inclusion, the non-bank has appointed Dr Jane Watts to its board. ...

The non-major bank has taken another leap towards the digital space, integrating its core platform to the cloud. ...


Join Australia's most informed brokers

Do you know which lenders are providing brokers and their customers with the best service?

Use this monthly data to make informed decisions about which lenders to use. Simply contribute to the survey and we'll send you the results directly to your inbox - completely free!

Do you think the new NSW property tax will help or hinder first home buyers?

Website Notifications

Get notifications in real-time for staying up to date with content that matters to you.