Westpac is looking to expand the use and efficacy of AI chatbots across the bank, and has invested millions of dollars into the US tech company that makes them work.
has turned technology investor, pouring money into New York-based conversational artificial intelligence software firm Kasisto. As a lead investor in a $US15.5 million funding, Westpac hopes to gain the inside track on future product development.
As well as chatbots, nine-year-old Kasisto also makes a so-called orchestration platform, which links up different bots from different vendors, and will move conversations around the bank to the most relevant ones, rather than ending a customer conversation when it runs out of answers. “The ability to orchestrate across chatbots as they become even more important – I think there’s about 4 billion of them around the world now – is going to be kind of a game-changer for us, as there are so many use cases out there now.”Alongside Westpac in the funding round were US-based Fidelity Information Services, and BankSouth, a US community bank in Georgia, which has also been an early adopter of Kasisto’s KAI platform.
It began working with Kasisto first in 2020 to improve its suite of chatbots, before launching a digital assistant called Blue last year to help advisers and direct investors using BT’s wealth management platform. During the COVID-19 pandemic, Westpac brought back to Australia over 1000 customer service jobs that had been conducted in India and the Philippines, after customers were left waiting too long for call centre responses and home loan processing.Asked whether the greater adoption of AI chatbots was a measure to save costs on more expensive customer service salaries in Australia, Mr Collary said this was not the intention of the plans.
Kasisto’s co-founder and chief executive Zor Gorelov said some chatbots deserved the cynicism with which they are viewed by customers, due to their limitations, but that the work it was doing with Westpac was effectively developing the next generation of digital assistants.He said this meant software that could predict and anticipate users’ needs, and that specialist chatbots worked better than the general ones that people often encounter on websites.