YAPA started from a simple frustration: both of us were binging great personal‑growth podcasts and retaining almost nothing. We could quote a handful of abstract concepts and we had no sense of what we actually understood. That gap (tons of high‑quality audio, almost no structure or assessment) became the seed for a learning layer on top of podcasts.
We approached it as a learning problem, not a summarization problem. From cognitive neuroscience and learning research, we knew that active recall, spaced repetition, and clear learning objectives drive long‑term retention. So the PRD focused on transforming podcast episodes into short, structured lessons with transcripts, concept extraction, quizzes, and a spaced‑repetition review queue, in addition to mastery indicators at concept and course level.
We built the project as a web app: first, a pipeline to accept a podcast URL, fetch or upload a transcript, segment it into meaningful chunks, and extract key concepts. Then we added an AI‑assisted layer to generate draft questions, which we manually reviewed and tagged to concepts and timestamps. On top of that, we implemented a lightweight quiz and SRS engine and dashboards showing lesson completion, accuracy over time, and concept mastery.
Log in or sign up for Devpost to join the conversation.