In this lightning talk, I share an innovative, human-centered workflow for transforming Open Educational Resources (OER) into dynamic, multimodal language-learning content using artificial-intelligence–generated avatars. By leveraging HeyYen’s AI-driven avatar technology—capable of producing natural body movement, facial expression, and authentic voice output—I converted a traditional OER Spanish textbook into a complete library of more than sixty short instructional videos. These micro-lessons (approximately two minutes each) deliver grammar explanations, vocabulary readings, pronunciation models, and cultural notes in a format that is highly accessible to diverse learners.

This project demonstrates how instructors can repurpose static text into engaging, culturally relevant media without requiring video-production expertise or expensive studio tools. The resulting videos provide consistent modeling of target-language input, reduce cognitive load through micro-learning design, and support students who benefit from repeated exposure, multimodal instruction, and flexible pacing. Importantly, the process foregrounds “human-by-design” principles: instructors maintain full control of content, sequencing, and pedagogical approach while using AI only to enhance clarity, engagement, and access.

Participants will gain insight into the workflow used to convert OER text into avatar-based micro-videos, including practical steps, design considerations, and examples of how these videos can supplement in-person, hybrid, or fully online language courses. The session highlights an affordable, scalable method for creating high-quality instructional media that expands accessibility and supports learner autonomy, especially in ESL and Spanish classrooms serving multilingual, multicultural students.