Article

HR Answers: Leveraging AI to Its Full Potential

digital illustration of robot projecting from laptop computer screen processing data
By Dan Michaeli

4 minutes

How credit unions are using artificial intelligence coaching for member service representatives

While many credit unions already leverage artificial intelligence for chatbots and handling simple member-facing questions and interactions, AI’s applications can—and should—be much broader. Innovative credit unions are incorporating AI into their member service strategy in new and expanded ways, including agent training and coaching or support, positioning them to more effectively serve their members. Utilizing this technology in such a manner leads to both increased efficiencies and reduced costs.

Leveraging AI to assist member service representatives can be a win-win for credit unions and members. AI-driven bots can surface relevant information to agents during member interactions, relieving them from having to retrieve data from backend systems and ultimately speeding up interactions. In addition, bots can scan chats with members in real time and present suggested answers or prompts, which can be accepted, discarded or modified by the service representative, reducing time and effort for employees while improving service for members.

Another key benefit of using AI coaching bots is their ability to oversee member interactions. The bots can monitor agents to make sure they are responding to members in a timely manner, help them overcome roadblocks and answer their questions correctly. This also can help keep everything running smoothly from a compliance, security and accuracy standpoint.

An example of a credit union leveraging AI in innovative, strategic ways can be found in PSECU, Harrisburg, Pennsylvania. The $7 billion credit union has been successfully leveraging Glia’s digital member service platform for years and recently started capitalizing on Glia’s AI management capabilities to deploy conversational bots as operator assistants as well as a member-facing virtual assistant named Walter.

PSECU has benefited from the platform’s AI coaching capabilities for service representatives. By giving employees access to AI-powered coaching tips, the credit union has reduced both its response time and overall duration of member chats.

“Once our agents became accustomed to having the coaching suggestions populate the screen, they quickly realized the benefits they provided,” says Brad Heiser, lead manager of transaction services contact center for PSECU. “The tips greatly reduce the amount of typing needed when responding to members’ questions, which in turn reduces handling time and allows the agents to more easily manage multiple chats.

“One of our agents’ favorite features is the ability to alter the responses and tailor them to more directly answer a member's inquiry,” Heiser adds. “The added benefit to this feature is that the system captures the edited response, allowing for review and possible updates or additions to the coaching responses. The agents are also encouraged to provide input on improvements they feel can be made to the coaching tips for use in live chats as well as enhancements to our chatbot responses.”

Allowing agents to tweak responses strengthens the bots’ intelligence by creating more data; data compiled from the accepted, declined or modified suggestions is used to improve recommendations in the future. As a result, the bots progressively become more reliable, eventually reducing or eliminating the need for the credit union’s team to manually refine suggested responses over time.

“We believe that AI coaching for representatives will assist us in reaching our ultimate goal of a 24/7 touchpoint for members in our digital channels,” says Heiser, “which should lead to reduced volume in other channels such as phone and email.”

Since debuting in late 2019, PSECU’s Walter interactions have steadily grown in volume and accuracy, handling nearly 75% of chats successfully and roughly half without human assistance. This can be attributed to the credit union’s careful strategic planning prior to launch and ongoing optimization efforts. Since social distancing hit, chats spiked to nearly double for transaction services. Walter helps PSECU effectively service members in need without requiring additional staff. Each AI-enabled self-service chat saves time for members and money for PSECU when multiplied over thousands of interactions that would average five to six minutes each with a live representative.

Leveraging agent-assisting bots to their full potential is an initiative that is only going to become more impactful and important for credit unions to prioritize in the coming months. In fact, we predict that more innovative financial institutions will adopt such technology during the last quarter of this year and into 2021. Finding a way to blend the human touch with AI should be a key focus for credit unions, especially in the wake of the pandemic. Using AI to assist service representatives will enable credit unions to improve their operations and positively contribute to their bottom lines while delivering a more efficient, seamless and convenient experience to members.

Dan Michaeli is CEO and co-founder of Glia, a provider of Digital Customer Service, with offices in New York and Estonia.

CUES Learning Portal