Math education remains a critical focus for national education improvement in the U.S. Given the benefits of personalized learning and affective interventions for math learning, schools and districts seek tools with such support to engage students. We report on recent findings from a large-scale randomized controlled trial of a technology-based learning platform that leverages AI-based adaptive technology to offer personalized content, remedial tutoring, immediate feedback, and affective support for students and provides learning analytics on student performance and effort for teachers. The experimental study involved 64 math teachers. Analyses suggest that implementing the program helped teachers adjust their practices, and students who used the program more performed better than the comparison group after controlling for prior performance.