What types of data exist in Horizons?

Last updated: July 30, 2025

Horizons integrates a wide range of publicly available information (PAI), curated by our analysts and trusted partners, to help uncover hidden connections and support complex investigations into illicit networks.

1. Data Source Types

Horizons hosts a diverse collection of data sources selected for their investigative value:

  • Corporate & Property Records – Corporate registries, beneficial ownership disclosures, and property filings that help identify individuals or entities behind companies and assets.

  • Trade & Customs Data – Import/export filings and shipping records that reveal supply chains and cross-border transactions.

  • Transportation & Asset Tracking – Aircraft and vessel movement datasets to monitor assets in near real time.

  • Screening & Watch Lists – Sanctions, politically exposed persons (PEP) lists, and other high-risk screening datasets.

All datasets are reviewed, standardized, and indexed to ensure reliability and easy access.

2. Data File Types

Horizons supports a variety of file formats:

  • Tabular Data: CSV, XLSX, and similar structured files

  • Documents & Reports: PDFs, HTML pages, and text-based records

  • JSON and API-derived data: Structured machine-readable outputs for more complex datasets.

These are combined into a single unified search experience so investigators can work across multiple formats seamlessly.

3. Modeled & Structured Data (Follow the Money Schema)

Horizons doesn’t just store raw data—it models complex datasets using the Follow the Money (FtM) schema, a standardized framework for describing entities and relationships.

The FtM schema organizes information into:

  • Entities (e.g., Company, Person, Address, Asset, Contract)

  • Relationships (e.g., ownership, directorship, shareholding, associated with)

For example, one corporate registry entry might list a company, multiple directors, shareholders, and related addresses. Using the FtM schema, Horizons breaks this single entry into:

  • A Company entity representing the business itself

  • Multiple Person entities for directors and shareholders

  • Ownership and Directorship relationships linking those individuals to the company

  • Associated Address or Asset entities

This modeling approach allows investigators to search, filter, and map out networks quickly, helping to reveal patterns such as hidden ownership structures or cross-border connections that would otherwise be buried in raw tabular data.