Globodex Specification
Event Definition
Standard
A shared language for describing what happens in the world: where, when, and why it matters. EDS gives AI agents and media systems a common way to discover, verify, and reason about real-world events.
Version V 0.1 — Working Draft
Scope Events · Entities · Media · Trust
Published By Channel 1
Learn more
EDS is a specification that describes events (things that happen at a specific time and place) and links them to verifiable media assets like video, audio, images, and text.
Designed for the Agentic Web
The Problem
Event data is fragmented,
unverifiable, and siloed
Every newsroom, sensor network, and AI agent produces data about what's happening in the world. But there's no shared way to describe it, and that creates real problems.
?
No shared schema
Every system uses its own format for describing events. Reuters and AP use different schemas. Sensor networks speak a different language entirely. Merging them requires expensive, brittle custom integrations.
!
Trust is an afterthought
Provenance, verification status, and rights metadata are rarely baked into the data model. They get bolted on later, or lost entirely. In an era of deepfakes and synthetic media, this is untenable.
×
Agents can't reason
AI agents need structured, machine-readable event data with spatial and temporal coordinates to answer questions like "What just happened in downtown LA?" Today, that data doesn't exist in a queryable form.

How It Works
One schema for everything
that happens

At its core, EDS defines a single portable object, the event, that captures what happened, where it happened, when it happened, and what evidence exists for it.

Every event carries temporal bounds (when it started and ended, with uncertainty margins), a spatial footprint (a GeoJSON polygon covering the area, plus altitude and H3 spatial indexing), and a classification that describes what type of event it is.

Attached to each event are the raw media assets (video clips, photos, audio recordings, text dispatches) each with their own provenance chain, cryptographic hashes for tamper detection, and optional cryptographic manifests for end-to-end content authenticity verification.

Events then connect to each other through typed, weighted relations, forming a knowledge graph over time and space. One event precedes another. A traffic disruption is caused by a march. Two reports from different sources are flagged as same_as, or contradicts.

The result is a structured, queryable, verifiable record of what's happening in the world, designed for both human editorial teams and autonomous AI agents.

Architecture
Four-layer data stack
EDS is designed to work with a four-tier storage architecture. Raw media flows in at the bottom and gets progressively enriched: extracted into metadata, embedded into vectors for semantic search, and finally linked into a knowledge graph.
OBJECT STORAGE VIDEO · AUDIO · IMAGES · TEXT · TELEMETRY L0 L1 L2 L3 METADATA STORAGE TIME · PLACE · CLASSIFICATION · PROVENANCE · TRUST VECTOR STORAGE EMBEDDINGS · SEMANTIC SEARCH · SIMILARITY GRAPH DATABASE ENTITIES · EVENTS · RELATIONS · KNOWLEDGE GRAPH EXTRACT EMBED ANALYZE ORIGINAL ASSETS RAW MEDIA FILES CONFORMANCE SCORING L0 MINIMAL → L3 VERIFIED L0: BASIC · L1: TIME+PLACE L2: FULL METADATA · L3: VERIFIED SEMANTIC EMBEDDINGS METADATA + OBJECTS → VECTORS KNOWLEDGE GRAPH ENTITIES · EVENTS · EDGES STORAGE ARCHITECTURE EXPLODED VIEW QUERY LAYER ↑ RETRIEVAL · REASONING INGEST LAYER ↑ ASSETS FLOW UPWARD
Storage architecture — Exploded view showing four data tiers with conformance levels and processing flow
Ingest layer
Object Storage
Raw media files enter here. Original assets are stored in their native format with full fidelity preserved.
Video Audio Images Text Telemetry
Extract
Enrichment layer
Metadata Storage
Structured metadata extracted and enriched by AI agents. Each record is scored with a conformance level indicating completeness and verification.
Time Place Classification Provenance Trust
Conformance levels
L0 L1 L2 L3
Embed
Search layer
Vector Storage
Metadata and objects are converted into vector embeddings, enabling semantic search and similarity matching across the entire corpus.
Embeddings Semantic search Similarity
Analyze
Query layer
Graph Database
The knowledge graph layer where entities, events, and relationships are linked. Agents retrieve and reason over structured intelligence here.
Entities Events Relations Knowledge graph

Object Storage holds the raw media files: video clips, photographs, audio recordings, sensor telemetry. These are the source-of-truth assets, stored immutably with cryptographic hashes.

Metadata Storage extracts structured information from those assets: when they were captured, where, by whom, with what device, and how much of the EDS schema they conform to. Conformance is scored on four levels, from L0 (basic, minimal fields) up to L3 (fully verified with cryptographic content authenticity).

Vector Storage generates semantic embeddings from both the metadata and the original media, enabling natural-language queries like "protests in downtown LA this week" to return relevant events by meaning, not just keyword match.

Graph Database is where events become knowledge. Entities (people, organizations, places) are linked to events through typed relations, forming a knowledge graph that supports causal reasoning, timeline reconstruction, and cross-source deduplication.


The Event Model
Anatomy of an event
The event object is the atomic unit of EDS. Every event is a structured record of something that happened, with six layers of information that give it context, verifiability, and meaning.
Event Object — Layer Architecture
Identity
schema · id · status · language · summary
Time
interval · lower · upper · uncertainty
Place
geometry · altitude · crs · spatial_index (H3)
Classification
type · subtype · taxonomy
Provenance
created_at · created_by · collection_method · confidence
Trust & Rights
verification · signals · license · consent · restrictions

Identity gives each event a globally unique, time-sortable identifier (UUIDv7 or ULID), a human-readable summary, and a status indicator (whether it's been observed, predicted, or retracted).

Time captures when the event occurred as an ISO 8601 interval with explicit upper and lower bounds. Critically, it also records uncertainty, because real-world events rarely have precise start and end times. An earthquake might be timed to the second; a political rally might have 15 minutes of uncertainty on either end.

Place defines the event's spatial footprint as a GeoJSON polygon, not just a point. A wildfire covers an area. A parade follows a route. EDS captures that geometry natively, with altitude ranges and H3 spatial indexing for fast geospatial queries.

Classification categorizes the event using a pluggable taxonomy system. The base taxonomy covers common event types (natural disaster, public protest, election, press conference), but organizations can extend it with domain-specific types.

Provenance records who created the event record, when, using what collection methods (satellite imagery, mobile uploads, wire service), and with what level of source confidence. This is the chain of custody for the data itself.

Trust & Rights is where EDS diverges from most data standards. Verification status, trust signals (cross-source corroboration, cryptographic manifests), licensing terms, consent records, and usage restrictions are all first-class fields, not optional metadata bolted on later.

Example
Here's what an EDS event looks like in practice, a simplified example describing a march in Los Angeles.
event.json — eds-0.1
{
  "schema": "eds-0.1",
  "id": "evt_01HZYR4E1K0A3N7A6J3C4KZ9B2",
  "summary": "Large march along Wilshire Blvd",
  "status": "observed",

  // When it happened — with uncertainty bounds
  "time": {
    "interval": "2025-08-12T12:05:00Z/2025-08-12T13:40:00Z",
    "uncertainty": { "lower": "PT5M", "upper": "PT2M" }
  },

  // Where it happened — as a GeoJSON polygon, not a point
  "place": {
    "geometry": { "type": "Polygon", "coordinates": [/* ... */] },
    "spatial_index": { "h3": { "res7": ["87283082affffff"] } }
  },

  // What type of event
  "classification": { "type": "public_protest", "subtype": "march" },

  // Who recorded it and how
  "provenance": {
    "created_by": "org:channel1",
    "collection_method": ["uav_video", "mobile_uploads"],
    "source_confidence": 0.86
  },

  // Trust and rights travel with the data
  "trust": {
    "verification_state": "partially_verified",
    "signals": [{ "kind": "cross_source", "score": 0.7 }]
  },
  "rights": { "license": "CC-BY-4.0", "restrictions": ["no_face_recognition"] }
}

Capabilities
What EDS enables
Beyond the core event model, EDS provides building blocks for media assets, knowledge graphs, content verification, and progressive data quality.
Media asset linking
Attach video, audio, images, and text to any event with full provenance: who captured it, when, with what device, and where. Each asset carries a cryptographic hash for tamper detection and an optional cryptographic manifest for end-to-end content authenticity.
Knowledge graph relations
Events connect through typed, weighted, directional relations. Temporal relations (precedes, follows), causal relations (causes, caused_by), structural relations (within, superset_of), and identity relations (same_as, contradicts) enable automated reasoning across sources.
Built-in trust and verification
Every event carries a verification state and an array of trust signals: cross-source corroboration scores, cryptographic verification, geolocation validation, temporal consistency checks. Trust isn't a separate system; it's embedded in the data model.
Rights and consent as first-class fields
License type, consent status (faces, voices), permitted usage categories, and machine-readable restrictions travel with every event. Downstream consumers can automatically enforce policies like "no face recognition" without human review.
Conformance Levels
Progressive data quality

Not every data source can provide every field on day one. EDS uses four conformance levels so that teams can start contributing data immediately and improve quality over time. An event at L0 is still useful; it just has fewer guarantees.

L0
Minimal
Schema version, event ID, and summary. Enough to register that something happened.
L1
Localized
Adds time and place. Now the event is anchored in the real world with temporal bounds and a spatial footprint.
L2
Complete
Full metadata including classification, provenance, rights, trust signals, and entity extraction. Production-ready.
L3
Verified
Cryptographic content authenticity manifests attached. Cryptographic proof of provenance from capture device to publication.
Relation Types
How events connect

EDS defines four categories of relations between events. Each relation is directional, carries a confidence weight, and can optionally link to the specific assets that evidence the relationship.

Temporal — sequence and co-occurrence
precedes follows coincides_with
Causal — cause, effect, and correlation
causes caused_by correlates_with
Structural — hierarchy and composition
within superset_of
Identity — deduplication and contradiction
same_as duplicate_of supersedes contradicts

Design Principles
Why EDS is built this way
Five principles guided the design of EDS, from the composable core schema to the extensions registry.
01
Small core, rich extensions
The base schema stays minimal and readable: six required fields that any system can produce. Advanced capabilities (custom taxonomies, domain-specific metadata, ML model outputs) come through a governed extensions registry, so the core never bloats.
02
Standards-native interoperability
EDS uses JSON natively, GeoJSON for spatial data, ISO 8601 for time intervals, and provides JSON-LD hooks for knowledge graph integration. It's designed to work alongside existing geospatial, media, and news standards, not replace them.
03
Verifiability by default
Provenance, cryptographic hashing, and content authenticity verification are built into the schema, not optional add-ons. Every event and every asset can be traced to its origin and verified against tampering, at every layer of the stack.
04
Safety and rights are first-class
Consent records, license terms, and sensitivity flags are required fields, not afterthoughts. Rights metadata travels with the data through every storage tier, every API response, every downstream consumer. Machine-readable restrictions enable automated policy enforcement.
05
Predictable evolution
Semantic versioning with explicit deprecation timelines and an extensions registry. Breaking changes are announced in advance with migration guides. Your integration won't break on a Tuesday morning.
Start building with EDS
EDS is a working-draft specification. We're building it in collaboration with broadcasters, publishers, and AI companies who need a shared language for event intelligence.