Michigan State University Federal Credit Union’s external virtual agent does the equivalent work of 60 full-time equivalent employees.
Financial institutions looking to succeed must provide customers with as accurate and timely service as possible. Beyond just maximizing efficiencies, keeping up with larger digital trends allows institutions to remain competitive, and equally important, their customers take notice. This focus is even more important to credit unions, heightened by their cooperative, member ownership structures.
Operating at the leading edge of digital transformation, Michigan State University Federal Credit Union (MSUFCU) consistently leverages its innovation lab to explore new ways to offer better technical experiences to their more than 345,000 members and over 1,100 employees. They’ve used boost.ai’s conversational AI to implement an internal virtual agent, Gene, as well as a member-facing virtual agent Fran.
MSUFCU's results at a glance...
- 98% resolution rate
- 620% increase in intent library
- Over 200,000 member conversations
Meeting the Measure of Member Demand
Member behavior doesn’t follow the nine-to-five schedule, nor does it consider the capacity of the service team. MSUFCU needed an external agent to increase their responsiveness both inside and outside of standard business hours, while ensuring accuracy in replies.
After an initial project with a previous vendor, MSUFCU selected boost.ai to optimize the virtual agent, rapidly scaling up a re-launch of Fran in September of 2021. Upon deployment, the Credit Union was able to provide 24/7 support for their members, expediting common questions, and ultimately providing more relevant responses and services for everyday demand. Within the first two weeks, Fran achieved an overall resolution rate of 81.1%.
Since then, MSUFCU's AI trainer team has continued to expand Fran's knowledge and optimize the virtual agent's ability to answer inquiries, increasing its resolution rate to 98%. This has dramatically improved the member experience and allows agents to focus on more nuanced support and service inquiries.
Part of the Team
Today, Fran is much more than just a virtual agent, but also a valuable member of the MSUFCU eServices team. Fran has been integrated into MSUFCU’s live chat service, being the first contact point before members chat with a live agent. The Credit Union began calculating how much work Fran is doing and discovered the virtual agent is handling the equivalent chat load of nearly 60 full-time employees, equating to almost 45,000 conversations per month.
With the virtual agent handling that level of volume, the eServices team can now provide a greater degree of care for the queries that require it.
“Unlike AI bots of the past, Fran can assist in simple and somewhat complex situations, sometimes fully assisting a member without needing an employee’s help at all. Fran takes a good chunk of the live chat workload, which also helps us have more in-depth conversations with members.”
-- Sam Kirkey, eServices Specialist
Looking Ahead
These initial successes haven't slowed MSUFCU down in looking to what’s next for Fran. With an eye towards the future, their team of AI trainers and technologists continue to refine Fran’s relationships with their members and employees.
Recently the team automated 10 common account processes through multiple API connections, allowing Fran to present personalized information to their members on rates, account balances, payments, loan payoffs, and more. They’ve also launched proactive chat functionality to optimize Fran’s position in their service workflow, allowing her to respond to member needs when and where they arise.
The team from MSUFCU has leveraged their expertise and member feedback to build out a savvy, responsive, and evolving member support in Fran. With more time to experiment and optimize Fran and their many other plans, they’ll no doubt continue to improve what is already one of the premier member experiences in the country.