Personal AI Tutor

Overview

PAT (Personal AI Tutor) is a human-centered learning companion designed to nurture curiosity, reflection, and agency in early education through conversational AI, computer vision, and immersive environments. The project explores how AI and XR systems can support learning without replacing human guidance — positioning technology as a guide rather than an authority.

PAT was developed as part of “Designing Virtual Worlds” at MIT, a course taught by Professor Ken Zolot, and brought together an interdisciplinary team of students from MIT, Harvard, and Berklee College of Music. The project was mentored by Alvin Wang Graylin, XR pioneer, author of Our Next Reality, and a leading voice in immersive computing and human-centered technology.

The system allows children to interact naturally through voice and image input, asking spontaneous questions about their environment and receiving age-appropriate, curiosity-driven responses. PAT adapts over time, generating daily reflection summaries for both children and parents — prioritizing transparency, safety, and trust while preserving the child’s sense of autonomy. A Unity-based plug-in enables PAT to inhabit virtual worlds, responding contextually to objects, spaces, and learning objectives inside XR environments.

For this project, I served as Project Manager and Audio Director, helping shape PAT’s product vision, interaction flow, and experiential coherence. I contributed directly to Unity scene creation, audio design and implementation, and XR integration, ensuring the system functioned reliably and intuitively on headset hardware. I also led quality assurance and play-testing, validating real-time voice, image capture, and conversational stability across multiple virtual environments.