LONDON — DNB’s internal chatbot fielded more than 2.1 million queries last year and boasts an 83% accuracy rate. 

Juno, the chatbot launched in March 2020, “knows how to answer around 3,500 questions and he can also answer them in seven different ways, depending on what kind of adviser is asking the question,” Jorgen Hansen, AI trainer at Oslo, Norway-based DNB, said today at the Artificial Intelligence in Financial Services 2023 conference in London. 

The AI chatbot, which the bank built with Boost.ai, was trained on millions of entries and responses on DNB’s SharePoint database, Maia Sognefest, product manager at DNB, said during the event. The data and responses are used by company representatives to resolve customer queries, she said.  

The representatives can even provide feedback when Juno’s responses are incorrect so that the tech team can improve the chatbot, Hansen said. 

Streamlined employee onboarding

Along with improving query resolution, the chatbot has allowed the bank to speed up employee onboarding, improving employee satisfaction, Sognefest said. 

“When you have new people coming in, it’s easier for them to learn the way we’re working when they have Juno instead of old SharePoint sites when it used to take a year to know where everything was. … You save so much time,” she said. 

Michigan State University Federal Credit Union has had similar success with Boost.ai. The $7.7 billion credit union has a 98% accuracy rate with its chatbot, Fran, and that has allowed the bank to free up human resources throughout its operations, Ami Iceman-Haueter, chief research and digital experience officer, previously told BAN. 

Join Bank Automation News for the upcoming webinar, Global Ideas for Better Banking AI, on Thursday, Sept. 14, at 11 a.m. ET.  For more information on this free webinar and to register, click here.