Roadmap

Botverse Metadata

OMC-based knowledge graph for media assets. Store files, entities, and their relationships. Query by character, scene, asset type, or any OMC property.

⚠

Botverse Metadata is on the roadmap. The schema and tool signatures below represent the design intent. They will be finalised before the service ships.

What it is

Botverse Metadata is a graph database service that uses the MovieLabs Ontology for Media Creation (OMC) as its descriptive layer. OMC is the industry standard for describing media production entities — used by major Hollywood studios and co-created by Botverse's founder.

The graph stores media assets as nodes (CreativeWork, Asset, Character, Scene, Depiction) and the relationships between them (derivedFrom, depicts, extractedFrom, transcribedFrom). An agent can then query: "find all footage of character X in outdoor scenes during acts 2-3 of the script."

Example: ingesting a pipeline output

ingest_pipeline_output — creating a full OMC graph

{
  "creative_work": {
    "title": "Episode 3 — Desert Chase",
    "omc_class": "CreativeWork"
  },
  "assets": [
    { "id": "source",     "omc_class": "Asset", "asset_type": "Video.Source",     "object_key": "uploads/..." },
    { "id": "mp4",        "omc_class": "Asset", "asset_type": "Video.Transcoded", "object_key": "outputs/..." },
    { "id": "thumbs",     "omc_class": "Asset", "asset_type": "Image.Thumbnail",  "object_keys": ["outputs/thumb-0.jpg", "..."] },
    { "id": "transcript", "omc_class": "Asset", "asset_type": "Audio.Transcript", "object_key": "outputs/transcript.vtt" }
  ],
  "relationships": [
    { "from": "mp4",        "type": "derivedFrom",    "to": "source" },
    { "from": "thumbs",     "type": "extractedFrom",  "to": "source" },
    { "from": "transcript", "type": "transcribedFrom","to": "source" }
  ]
}

Planned tools: ingest_pipeline_output, find_assets, get_asset, query_graph, link_entities, update_asset.