{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/bdf29c008b244605b2b7efd4ddbfc1b8\" frameborder=\"0\" width=\"1280\" height=\"960\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":960,"width":1280,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":960,"thumbnail_width":1280,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/bdf29c008b244605b2b7efd4ddbfc1b8-9d592125af0cf423.gif","duration":223.765,"title":"NexyClaw Converts Interviews into Knowledge Graphs","description":"This Loom presents NexyClaw for TrackOne, a reflective ingestion gateway that turns human interview experiences into structured knowledge for Atomspace. The speaker addresses the VGI challenge of “flattening,” where standard LLMs simplify nuanced, culturally grounded context into generic business language that logic systems cannot use. NexyClaw uses a two-stage approach: one LLM filters generic speech and identifies real cultural context, tensions, and relationships, and a second LLM converts those into validated meta-expressions for knowledge graphs. It integrates into Omegakla as a modular plugin skill with token orchestration, storing outputs directly into Omegakla’s local knowledge graph, and includes a live URL demo deployment.\n\nURL: https://2e438d04a655639727.gradio.live/"}