Banking giant TD has expanded its AI-backed chatbot platform into the U.S. to better assist its customers during the pandemic. The AI tech is being integrated with TD’s existing virtual assistant, “Clari,” to provide better insights and hold more conversations with customers. 

“We saw the traffic on Clari double since the pandemic,” said Imran Khan, head of global digital experience at TD, during a webinar hosted Thursday by Bank Innovation. “We wanted to win in customer engagement, so we are looking for ways to improve our tech platform.” 

The feature, which runs on technology from the conversational AI company Kasisto, has accelerated TD’s  technological roadmap, Khan said. In addition to Kasisto’s “deep learning AI tech,” the platform integrates TD’s data to give better customer insights, which are used with interactions data to improve its AI platform, he added. 

“In the first week of launch, the AI tool supported 75,000 customers sessions with over 135,000 unique interactions,” Khan said. TD has $1.1 trillion in assets and 14 million customers worldwide.  

Because the pandemic has forced customers to use digital channels, TD plans to add new features to its tech platform to maintain that online channel engagement. 

“People are trying to be more self-reliant through different technological platforms,” Khan said. “We want to evolve as well, to meet with consumer needs and demands of the current times.” 

TD plans to partner with different fintech companies to improve its platform and “create better experiences that drive engagement, through technology that’s already been built by other fintechs,” Khan added. 

The Toronto-based bank also has plans this year to launch a project it is calling internally “MyTD” along with Layer6, an AI machine learning company TD acquired in 2018.   

Layer6’s predictive AI tool tells customers where their balance might go in the coming weeks or whether they can pay any upcoming bills, Khan explained. 

“When the pandemic hit, we realized that providing our customers with such tools and insights has never been more important,” Khan said. “We wanted to give our customers more than traditional AI capabilities like an expense predictor.”