This talk introduces PING, a scalable, AI-powered, web-based education platform designed to transform how students engage with learning. PING delivers modular, grade- and topic-based learning experiences centered on high-impact instructional pain points identified by practicing teachers. These modules are implemented as interactive, gamified activities that generate rich learning signals beyond simple right-or-wrong responses, including misconception patterns, strategy choices, and engagement trajectories. Embedded within these experiences are two intelligent agents. The Peer Agent questions, challenges, and reasons with learners, prompting them to explain, justify, and teach concepts back to the system, thereby fostering metacognitive reflection and learning-by-teaching behaviors. Operating alongside it, the Tutor Agent provides adaptive academic scaffolding in real time. Together, these agents address a key limitation of traditional digital learning systems: they support not only what students learn, but how they think about their learning — at scale.
