Enterprise content is often scattered across folders, Slack threads, decks, drafts, and old client work. That makes it hard for AI agents to know what is current, approved, and safe to use. This demo shows how Box can help teams turn scattered content into trusted AI knowledge using Box Hubs, Box Automate, and MCP. A new Retail discovery asset is uploaded into a staging folder, reviewed by an AI agent through Box MCP, enriched with proposed metadata, routed through Box Automate for approval, and published into a curated Box Hub. Once approved, the content becomes part of a governed knowledge base that AI agents can safely use to answer questions with trusted sources. Timestamps: 00:00 — The problem: scattered enterprise knowledge 00:15 — Organizing client knowledge by business domain 00:29 — Uploading a new asset into Retail staging 00:44 — Using Box MCP to review and summarize the asset 01:00 — Extracting proposed metadata, tags, domain, and health score 01:12 — Routing the asset through Box Automate for approval 01:32 — Publishing approved content into a Box Hub 01:48 — Asking a question and getting an answer from trusted knowledge 02:00 — Final takeaway: governed content as the knowledge layer for AI agents Featured: Box Hubs - https://www.box.com/hubs Box Automate - https://www.box.com/automate Box MCP - https://www.box.com/mcp-server Learn how Box helps teams build trusted, governed knowledge bases for AI agents, so agents answer from approved content instead of random files, drafts, and outdated documents.

 59      0