By W.B. King
More than 1,000 banking and fintech executives descended on San Francisco in late May to attend the FinovateSpring conference, which marked its first in-person event since 2019.
Among topics discussed was “Scaling Conversational AI – Going from Idea to Scaled, Seamless Experiences for Your Customer,” hosted by Laura Sievert, investor at CMFG Ventures, the corporate venture capital arm of CUNA Mutual Group.
“When we think about the future of financial services, we are incredibly invested in what that means for partnerships between traditional financial institutions, credit unions and community banks,” Sievert said, noting that CMFG Ventures has partnerships with approximately 95% of U.S.-based credit unions.
“We are continually making investments…we launched our fund about five year ago and have about 40 companies in our portfolio, including Posh, and when we talk to credit unions, what is top of mind for them is digitization, personalization, retaining their members and creating more loyal members — that’s where chatbots come in,” she continued. “We have been having conversations with players in this space for several years now.”
Sievert was joined on the panel by Posh Technologies Co-Founder and CEO Karan Kashyap. He explained that the Boston-based conversational artificial intelligence (AI) startup was a spinoff of lab work conducted at the Massachusetts Institute of Technology where he and his partner were engaged in graduate work.
“We saw an opportunity to bring that technology to financial institutions,” Kashyap said, adding that company was founded three years ago. “We are really focused on verticalized AI conversations, specifically on digital and voice channels from a context of customer service and user experience.”
Sievert said that in her experience some credit unions look to this technology to increase personalization and efficiency. Other credit unions have trouble staffing call centers and need help in this area. She asked Kashyap how successful conversational AI programs can be rolled out.
Noting that there are different value propositions with conversational AI, Kashyap said the most common approach is automation. The goal is a 24/7, seamless avenue for end users to automate transactions across digital channels.
“One interesting thing about AI that people often discount is the value of the data that is unlocked,” he said. “In a call center AI scenario, for instance, the AI [bot] is getting the firsthand experience to understand a member’s frustration, pain points and what information they are seeking.”
He continued. “Mid-sized financial institutions are usually not doing a whole lot with it [data]. They are getting a lot of calls, they are getting a lot of chats but the depth of understanding of what is actually in that data is lacking.”
Commenting on how Posh Technologies assists credit unions in unlocking these data points, Kashyap noted, “We have tools that will essentially self-diagnosis what we called coverage gaps…patterns of interactions that has caused the bot to be stumped or escalates to an agent.”
Does AI Mean Losing the Personal Touch?
A concern for many executives, Sievert offered, is that AI can diminish the trait credit unions are best known for: member trust gained from personal relationship building.
Kashyap conceded that it is a “valid” concern, noting that the AI and robotics movement can be viewed as removing the “human factor.” He noted that credit unions take a pride in the “human touch” and see it as a differentiator.
He reflected on the rise of ATMs in the 1970s and 1980s. Many banks and credit unions were resistant because it would cause customers and members to stop visiting branch locations. This, in turn, would hamper the ability to build relationships and upsell services.
“The initial hesitation to ATMs was very similar to the initial hesitation to digital banking…no one is going to use it. You lose the personal touch,” Kashyap said. “It’s just a different framework of mindset. ATM…many people take it for granted now…it solved for easy transactions, 24/7 availability. If you think about technology as a way to help your customer to have a better experience and reduce friction, I think that is what we are really focused on with AI. It not a magical bean…it’s a tool we can use to solve problems and make member and customer experiences better.”
How to Successfully Scale to Chatbot
Among benefits of an AI/chatbot platform is the ability to free employees to engage members on services that can be profitable, such as loans. To this end, Sievert offered that automation actually allows for credit unions to offer more personal touch interactions where it counts most. The problem, however, is that many credit unions she speaks with aren’t happy with the chatbot solutions they are currently running. She asked Kashyap what it takes to successfully scale a chatbot solution.
“We see the best FI launches when the FI sees this as a tool. It’s not this all-encompassing, omniscient AI system. We found that FIs that are very transparent that say…this is actually an AI meant to provide easy access to information, meant to be 24/7 and meant to help people get to the right agent and keep investing in it overtime…are successful.”
The goal, he added, is to always set benchmarks and determine if the AI is actually solving user problems.
“Some level of marketing is important to educate the population on what’s coming next,” he said. “The best launches are ones that there is a practice around the data and a focus on consistent improvement. Launch the product, see how people are using it and use that information to make it better.”
If you enjoyed this article, you might like reading these Finopotamus articles as well: