{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/219bf4ea14e849dd81a5ed1c45c57000\" frameborder=\"0\" width=\"1920\" height=\"1440\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1440,"width":1920,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1440,"thumbnail_width":1920,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/219bf4ea14e849dd81a5ed1c45c57000-f9897c96b911c368.gif","duration":135.36,"title":"Synapse Creates Evidence-Based Organizational Memory","description":"This Loom explains Synapse, a local first cognitive memory layer that turns workplace writing into episodic memories, evidence-based knowledge graphs, and reusable agent workflows. In the awake view, it imports 40 high-signal conversations from a real Slack export, groups threads, excludes bot and system noise, redacts sensitive fields, and uses local gamma inference through Cactus to classify episodes, storing them in Postgres and indexing with mem0 for semantic retrieval. In the sleep view, System 2 consolidates those conversations into an auditable knowledge graph with edges linked back to the Slack sources. Retrieval is episodic and uses prior episodes first, then writes improved answers back as new episodes, while it also observes successful recall workflows to propose versioned skills for human review before enabling agents via an MCP."}