The Infrastructure Layer for Economic Intelligence.

RedGraphs maps how companies connect, transact, and influence each other — across 600,000+ public and private entities worldwide.

600K+ Companies Mapped
2005–Present Historical Coverage
10 Relationship Types
S&P Global Distribution Partner
Economic Structure

The Global Economy Is a Network.

Markets operate through supplier relationships, subsidiaries, capital flows, and strategic dependencies. These connections determine who bears risk, who captures value, and how disruptions propagate. A network-based intelligence layer is structurally inevitable.

Distributed Across Jurisdictions

Business relationships are disclosed across millions of filings, spanning jurisdictions, filing regimes, and decades of corporate history. Structure requires extraction at scale.

Multi-Tier by Nature

Second- and third-tier dependencies define exposure. Supply chains, ownership structures, and capital flows extend well beyond direct counterparties.

Quantifiable When Structured

Concentration risk, counterparty exposure, and systemic dependencies become measurable when relationships are extracted, validated, and modeled as a temporal graph.

How It Works

From Primary Evidence to Network Intelligence

Every relationship in the graph originates from a primary source document, passes through structured extraction and entity resolution, and is placed in the broader economic network.

01

Extract

Regulatory filings, financial statements, and corporate disclosures are processed to identify relationship evidence. Each extraction traces to a specific sentence in a specific document.

02

Structure

Each extracted relationship becomes a structured edge: two resolved entities, a relationship type, temporal bounds, and a provenance chain linking back to source evidence.

03

Network

The verified relationship is added to the economic graph, connecting to the entity's existing neighborhood. Multi-tier dependencies and exposure paths become visible.

Governance

Built with AI. Verified by Evidence.

Designed for institutional decisions where provenance and accuracy are non-negotiable.

Every relationship traces to a specific sentence in a specific document. Automated extraction is verified by human reviewers. Corrections are non-destructive overlays with full audit history.

  • Evidence-Backed Relationships Source, page, and evidence sentence recorded for every edge
  • Confidence Scoring Extraction quality, entity resolution, and temporal freshness scored
  • Human Validation Purpose-built review workflow with correction overlays and audit controls
{
  "relationship_id": "rel_8f2a1c",
  "entities": [
    "AAPL",
    "TSMC"
  ],
  "type": "supplier_customer",
  "evidence": {
    "source": "10-K FY2024",
    "page": 47,
    "sentence": "Taiwan Semiconductor Manufacturing..."
  },
  "confidence": 0.94,
  "extraction": "automated",
  "reviewer": "analyst_12",
  "decision": "accepted",
  "audit_trail": [
    {
      "action": "extracted",
      "timestamp": "2024-03-15T09:22:00Z"
    },
    {
      "action": "reviewed",
      "timestamp": "2024-03-15T14:08:00Z"
    },
    {
      "action": "published",
      "timestamp": "2024-03-15T14:08:01Z"
    }
  ]
}

Ready to explore the economic graph?

Request access to the platform, data documentation, or a technical walkthrough.