An AI-assisted solution to creating authentic student mentorship connections.
Graditude is a non-profit that helps student leaders facilitate mentorship programs to ensure equal access to career guidance and growth. I designed a 0-1 feature that utilizes AI powered workflows to help leaders connect student mentees with mentors through a personalized matching tool and match database.
My Impact
60% increase in user satisfaction between version 1 and version 2 of this screen. Students across the country will be using this upon Graditude's launch in Summer 2026.
Most student mentorship programs lack structure and personalization.
We interviewed a focus group of 7 university student leaders to understand existing gaps in facilitating mentorship on college campuses, and found three main frustrations.
Misaligned connections
Students couldn't find mentors that matched their personal needs, which led to mentorship relationships fizzling out almost immediately after the first meetup.
No structured mentorship tools
Many student mentorship organizers would be keeping track of mentorship relationships via disconnected email threads.
Generic platforms are not university specific
While LinkedIn and ADPList are extremely useful platforms, they lack university context.
Main Takeaway
Campus org leaders struggle to contact alumni from lack of formal alumni programs. They need a platform that helps them match every student in the org.
We began our process by designing a lo-fidelity matching flow.
After our first meeting with Graditude stakeholders and developers, we received one crucial piece of feedback that changed the entire feature vision.
We needed to make our design scalable to 150+ users, so we came up with three separate features to expand our solution. These features consisted of an AI matching flow, a management database for admins, and a manual flow.
AI-powered matching flow
In order to make it easier on admins to match 150+ students, we decided that an AI match suggestion engine would be very helpful in allowing the admin to bulk match mentors/mentees. Instead of having to decipher compatible meetings times/career interests, the AI simply does it for the admin, and provides suggestions. The admin still has power to choose whether or not to accept these suggestions.
The Match Management Database
As we iterated on the automated matching flow, we realized there would be instances when users don't accept a match and need to be rematched. There needed to be a place for admins to track match statuses, so we designed a centralized Program Members dashboard that tracks every participant's status in real time.
Manual Matching
We kept the manual flow to allow admins to have the choice of making a manual match if they prefer.
Making Further Refinements
We decided that we could use the AI automated flow to provide swap suggestions to admin, so if an admin decides to delete a mentee from a mentor's stack, they get provided with suggestions of who they can replace that deleted mentee with.
Final Flow
