Learning Engineering Virtual Institute (LEVI) Partnership and Engagement
As part of the LEVI project, J-PAL is partnering with Carnegie Mellon University to evaluate their hybrid human-AI tutoring system, which provides customized tutoring to students based on their individual needs.
Carnegie Mellon University
J-PAL North America is proud to partner with the Learning Engineering Virtual Institute (LEVI), an ambitious initiative with the goal of doubling the rate of middle school math learning within five years, particularly for students from low-income backgrounds.
LEVI brings together interdisciplinary teams from seven leading organizations: Carnegie Learning, Carnegie Mellon University, Eedi, Khan Academy, Rising Academies, the University of Colorado Boulder, and the University of Florida. These teams are developing and testing innovative learning technologies that harness the power of artificial intelligence to transform tutoring, teacher training, and student support systems around the world.
Within LEVI, different “hubs” act as support teams with specialized expertise—such as curriculum design, AI development, or implementation strategy—to strengthen different aspects of program development. As the Evaluation Hub for LEVI, J-PAL North America plays a central role in ensuring that these innovations are rigorously tested and grounded in evidence. Through technical assistance, tailored guidance, LEVI-wide workshops, and connections with researchers within J-PAL’s network, J-PAL supports each team in designing and implementing high-quality randomized evaluations that assess program effectiveness, cost-efficiency, and scalability. Together with LEVI, we are also interested in understanding the potential for these programs to scale effectively once they demonstrate positive impact. By applying J-PAL’s proven methodologies, the Evaluation Hub strengthens the evidence base around AI-driven education tools and accelerates the learning cycle between research, implementation, and policy.
Project History
J-PAL North America joined LEVI as the Evaluation Hub in September 2023. In the first year, J-PAL began providing individualized support to LEVI teams and laying the groundwork for rigorous evaluations.
In subsequent years, J-PAL deepened this support by leading training sessions, connecting teams with J-PAL network researchers, offering pilot funding opportunities, and assisting with evaluation design and recruitment. By Year 4, J-PAL helped teams finalize full-RCT designs, review proposals, and begin recruitment for their studies.
Projects
Bridging the Gap: Evaluating AI-Powered Math Tutoring to Support Low-Income Middle School Students
Location:
United Kingdom
It Takes a Village (Plus AI): A Randomized Controlled Trial of Hybrid Human-AI Math Tutoring at Scale
Location:
United States of America
The Effect of AI-Enhanced Instructional Coaching on Student Outcomes
Location:
United States of America
Personalizing Math Instruction At Scale: A randomized controlled trial of MATHstream
Location:
United States of America
A Large RCT of ALTER-Math – AI Powered Learning by Teachers to Enhance and Renovate Middle School Math Learning
Location:
United States of America
Project Leadership
- Catherine Darrow
- Amanda Lee
- Caleb Smith
Evaluation Support
- Laura Ruiz-Gaona
- Jessicca Sashihara
- Lisa Turley-Smith
- Amanda Lee
- Sarah Sahni
- Jesse Gubb
- Caleb Smith
- Daniela Gómez Treviño
Funding Partner
Further Information
If you have any questions or want to learn more about our work with LEVI, please reach out to Amanda Lee [email protected].