MSU Federal Credit Union
Fran virtual assistant for members
A Michigan credit union's virtual agent grew to handle up to 75% of member chats, the work of about 55 staff
Problem
MSU Federal Credit Union serves about 330,000 members. When online demand jumped in spring 2020, live chats rose 130% in a month, and simple questions such as routing numbers, loan payments and digital banking problems took up staff time.
Approach
The credit union launched Fran, an AI virtual assistant on the boost.ai platform named after its first manager, in 2019 with about 250 intents, and relaunched it in September 2021. A dedicated in-house AI team keeps training it; by early 2025 it covered about 4,000 topics, runs 24/7 in English and Spanish, and hands members to a human representative when needed.
Results
Takeaway
An internal team maintained the system as it grew from about 250 intents to about 4,000 over six years. All figures came from the vendor, which said no one was laid off and used its own resolution measure.
Categories
- Industry
- Finance and insurance
- Business function
- Customer service
- Innovation type
- Process innovation
- AI technology
- Conversational and voice AI
- Adoption
- Vendor solution
- Evidence
- Vendor's case study
Sources
Source summaries
boost.ai reports that Fran handles 75% of member conversations, has a 98% resolution rate up from 81.1% at its 2021 relaunch, and does work equal to about 55 full-time staff without layoffs. It traces growth from about 650 intents at the end of 2022 to about 4,000 in early 2025 and more than 40,000 messages in January and February 2025.
MarTech Series reports boost.ai's 2022 announcement that Fran launched in 2019 with 250 intents and had grown 400% to over 1,100, serving about 330,000 members around the clock with a resolution rate above 96%. It notes live chats jumped 130% month over month in spring 2020 and quotes the credit union's digital strategy vice president on meeting that demand.
Full details are in the original sources.