New Awards Totaling More Than $3 Million Launch Connected Futures Across Four UMass Campuses and Power the Next Generation of MentorAI

In a rare convergence, three major funding commitments arrived within weeks of one another this summer, together totaling more than $3 million. The Axim Collaborative, an anonymous foundation awarded grants to the Center for Evidence-Based Mentoring at the University of Massachusetts Boston to conduct a rigorous evaluation of MentorPRO and MentorA across four campuses. And, the National Science Foundation issued a $1,051,777 Small Business Innovation Research Phase II award of over 1MM to MentorPRO to continue engineering of the MentorPRO platform and its AI co-pilot MentorAI. Together these investments help launch Connected Futures, a system-wide peer mentoring initiative that will reach every community college transfer student and every incoming first-year student at UMass Amherst, UMass Boston, UMass Dartmouth, and UMass Lowell by its third year, approximately 15,700 students supported by roughly 157 trained peer mentors. The initiative is led by Principal Investigator Jean E. Rhodes, Frank L. Boyden Professor of Psychology and Director of the Center for Evidence-Based Mentoring, with Co-Principal Investigator Elizabeth Raposa, Associate Professor of Psychology at Fordham University. We asked Professor Rhodes about the new awards and what comes next.

What are the aims of the projects?

The initiative responds to a problem that free college programs have made newly urgent. Massachusetts has produced double digit enrollment increases at community colleges through MassEducate, yet access alone does not guarantee completion. Nationally, only about one third of community college students ever transfer to a four-year institution, and fewer than half of those who transfer earn a bachelor’s degree within six years. At UMass Boston, nearly 30 percent of first-year students do not return for a second year. Students often leave because they lack the relationships and navigational support that help them make sense of complex institutions. Connected Futures addresses this gap by pairing each student with a trained peer mentor who initiates contact before the semester begins and maintains structured outreach across academic, financial, social, and well-being domains throughout the year. Mentors work within MentorPRO, an evidence-based platform developed with NSF support that grew directly from the Center’s research, and every student is offered a validated asynchronous adaptation of Connected Scholars, the Center’s social capital curriculum that teaches students to identify, initiate, and cultivate relationships with mentors and other sources of support. The foundation awards fund the Center’s research and implementation work, while the NSF award separately supports the platform engineering that the initiative relies on.

What makes MentorAI innovative?

At a moment when many institutions are adopting student-facing chatbots as a low-cost substitute for human relationships, MentorAI takes the opposite approach. It is a scientific co-pilot that works exclusively for human mentors, organizing mentee information into daily snapshots and surfacing campus resources and evidence-based strategies drawn from decades of mentoring research, all without ever communicating directly with students. This human at the helm design reflects the Center’s findings that students prefer AI that works in the background while a person they trust remains in charge. A recent peer-reviewed evaluation with more than 2,500 first-year students found that greater engagement with MentorPRO was associated with higher end-of-year GPA, academic self-efficacy, sense of belonging, and overall well-being, and at UMass Boston messaging between students and their mentors was associated with significantly increased odds of first-year retention. The NSF Phase II award funds the next generation of the platform, including intelligent matching, flash mentoring, dynamic knowledge updates, and automated resource curation. Just as important, Connected Futures embeds a two-level randomized evaluation. A system-wide randomized controlled trial will test whether AI-supported peer mentoring improves retention, academic performance, and help-seeking relative to business as usual, and a nested trial will randomly assign mentors to use MentorPRO with or without MentorAI. To our knowledge this will be the first rigorous causal test of whether an AI co-pilot makes mentors more effective and large caseloads more sustainable at the scale of a public university system.

What are your next steps?

The team is completing a very busy planning and implementation period that includes executing memoranda of understanding with each campus, naming campus coordinator leads at UMass Amherst, UMass Dartmouth, and UMass Lowell, and designing privacy-compliant data pipelines that link mentoring records with registrar data across four distinct student information systems. In Year One, thirty peer mentors will serve the incoming community college transfer cohort at a caseload of approximately 33 students per mentor, and the model will scale to first-year students across all four campuses in Years Two and Three. The two-year NSF Phase II period began August 1, 2026, and engineering milestones are already underway.

What do you hope will come out of this funded research?

We hope Connected Futures will demonstrate that proactive, relationship-centered support can be delivered affordably to every student in a broad access public university system, and that AI can make this possible by supporting mentors rather than replacing them. And because the model is designed, tested, and priced across a major public system, it offers a ready blueprint for other state systems absorbing growing transfer pipelines through free college programs. Sustainability is built in from the outset, with campuses positioned to absorb licensing and coordinator roles into standard operating budgets once the grant period ends.

Rhodes shared about the new awards. “For more than two decades our research has shown that caring relationships are the active ingredient in student success. What has been missing is a way to bring that ingredient to every student who needs it. These awards let us test whether AI, when it is designed to support mentors rather than replace them, can finally close that gap across an entire public university system. The convergence of foundation, system, and federal support tells us that funders across sectors see the same opportunity we do.”

Follow the Center for Evidence-Based Mentoring at cebmentoring.org, the Chronicle of Evidence-Based Mentoring at evidencebasedmentoring.org, and MentorPRO at mentorpro.com.