Structuring the Video Internet: How Vpin Uses Neo4j to Turn Watching into Learning

Session Track: App Dev

Session Time: 05:30 am - 06:00 am November 6

Session description

In this session, Chris Wei will begin by drawing your attention to a pressing issue: While the internet is saturated with video content, extracting and reusing actionable knowledge from these videos remains a significant challenge. Most video learning will continue to be passive and unstructured unless addressed. He will present Vpin’s mission: to transform passive video consumption into structured, searchable learning experiences. Wei will introduce Vpin as a video-first AI learning platform, powered by generative AI and Neo4j. You will see how users will curate videos, after which AI will extract key points and Neo4j will organize this information into a structured, reusable graph. This approach will support microlearning, expert knowledge capture, and streamlined team training. The session will then explain why Neo4j was chosen. Graph databases will enable the representation of complex relationships—between videos, highlights, concepts, tags, and users—facilitating semantic navigation and thematic exploration. Wei will walk you through the platform’s schema, showing how nodes and edges will map these connections. Sample Cypher queries will demonstrate how you will be able to quickly surface related highlights or concepts across multiple videos. The session will conclude with practical lessons: While graph search will excel at knowledge navigation, scaling and access control in multi-user environments will require careful management. Wei will emphasize that Neo4j’s flexibility, speed, and robust tooling will be central to Vpin’s mission. Ultimately, you will discover how Neo4j will transform scattered video content into a dynamic, navigable learning graph, with Graph RAG representing the next frontier.