Connected Metadata
How metadata enrichment and Super Tags work.
Automatic Enrichment
Every object that enters the system gets metadata on ingestion.
Upload a video and Base extracts speakers, topics, sentiment, rights holders, visual content tags, and temporal markers automatically. Drop in a brand guide and it pulls tone rules, color values, approved terminology, and usage restrictions.
This metadata is not static. It deepens over time as agents interact with the object. Usage history, relationship mappings, approval status, and compliance checks are added with every workflow execution. The richer the metadata, the smarter every agent query becomes.
Super Tags
Metadata is structured through Super Tags. A single asset can carry tags for:
- Rights status
- Content type
- Project association
- Agent ownership
- Campaign context
- Enrichment depth
- NILP® talent references
- Compliance status
Everything is searchable and queryable from a single layer. Super Tags replace the need for separate tagging systems, spreadsheets, or manual metadata entry.
How Agents Use Metadata
Agents do not search raw files. They query structured, enriched metadata records that already know what they contain. When a Brand Voice Agent checks an asset, it reads the tone tags, brand association, and prior approval history from the metadata layer. When a Compliance Agent reviews an output, it reads rights status, NILP® references, and jurisdiction flags.