// ZDZDATAVERSE · INTELLIGENCE ON THE BRAZILIAN ECONOMIC FABRIC

All of Brazil,
in one graph.

A canonical data lake that unifies dozens of official public sources into a single knowledge graph — with field-level provenance and LGPD by design. From one CNPJ you get the full 360°: company, establishments, ownership structure, sanctions, outstanding federal debt, trademarks. Built for risk, compliance, market intelligence and due diligence.

1.31 billion
consolidated rows — not a sample, but all of Brazil in public data
61.5 million
companies with ownership structures and chains of corporate control
27 sources
official public sources cross-linked in a single canonical model
686
tables in the canonical model
320 GB
of consolidated public data
64.5M
establishments (CNPJ)
24.7M
ownership relationships
LGPD
masked CPF · legal basis · no individual records
127M
external identifiers resolved
197M
provenance records
9
domains: corporate to environmental
// FROM ONE CNPJ TO 360°

One identifier goes in. The graph returns the entire context.

At the core is an entity graph connected by natural keys — CNPJ, IBGE code, CNAE, masked CPF. Starting from a single CNPJ, the path traverses establishments, partners, sanctions, outstanding debt and trademarks, resolving the same entity seen across different sources into one canonical record.

🏢
Input · CNPJ
00.000.000/0001-00 illustrative example
📍
Establishments
Headquarters + 4 branches illustrative example
🤝
Ownership structure
PARTNER·A · PARTNER·B illustrative example
⚖️
Sanctions / restrictions
Checked against official sources
💰
Outstanding debt · PGFN
Cross-referenced tax standing
™️
Trademarks · INPI
Industrial property portfolio
// values marked "illustrative example" are fictitious — no real data is exposed

Graph golden record

One company = one CNPJ with multiple roles. Heterogeneous sources become a single entity, with deduplication and survivorship — instead of silos that never talk to each other.

127 million bridges

External identifiers resolve the same company or person seen across 27 sources into one canonical entity. 360° enrichment from a single identifier.

Cross-references only a graph makes possible

Active companies in a sector within a municipality; a product's chain back to the raw material; ownership structures connected to sanctions and tax restrictions — all through natural keys.

🧭

Honesty about where the AI is. Entity resolution and matching in production today are deterministic and auditable (combination hashing, attribute similarity, survivorship). The generative-AI insights layer is roadmap, in rollout — and when it lands, it records its own provenance and never overwrites official data. AI + People: the machine accelerates, the evidence decides.

// EVERY FIELD IS AUDITABLE

197 million origin stamps. Traceable back to the official source.

Every piece of data in ZDZDataVerse carries provenance: which source it came from, when it was collected and with what confidence level. It is not a loose number — it is a number you can audit all the way back to the public document that originated it.

🏷️

Origin, date and confidence on every field

197 million provenance records stamp every value — the official source, the moment of collection and the assigned confidence level.

🛡️

Official data is never blindly overwritten

Survivorship with a golden rule: inferred data never erases official data. It coexists, flagged as inferred, with a complete trail.

🔒

LGPD by design

Personal data comes in with masked CPF and a recorded legal basis. The platform exposes scale and capability — never an individual person's record.

🔁

Always re-derivable

Idempotent, scheduled ingestion over named sources: the analytical table can be rebuilt from the raw source at any time.

fieldrazao_social
sourceReceita Federal · CNPJ
collected_at2026-07-12
confidenceOFFICIAL
fieldcnae_secundario
sourcecross-source enrichment
confidenceINFERRED
// provenance schema — illustrative values. Every real row points to the official source of origin.
// THE 27 OFFICIAL SOURCES

From Receita Federal's CNPJ registry to MapBiomas deforestation data.

27 official public sources, organized into 9 domains, cross-linked by natural keys in a single canonical model. From corporate to environmental, from tax to electoral — the Brazilian economic fabric seen from every angle.

🏢
Corporate

Receita Federal · CNPJ

Companies, establishments and ownership structures

📈
Finance · Markets

CVM

Listed companies, funds and securities

🏦
Banking

Estban

Banking statistics by municipality

💰
Tax

PGFN

Federal outstanding debt

🏛️
Public finance

Siconfi · Tesouro

State and municipal accounts

🗳️
Electoral

TSE

Candidacies, donations and campaign finance reports

™️
Industrial property

INPI

Trademarks and registrations

🌳
Environmental

MapBiomas

Land use and deforestation

🎓
Education

Censo Escolar

School networks and enrollments

🏥
Health

SIH · SUS

Hospital admissions

🛡️
Insurance

SUSEP

Insurance market and supervision

Energy

ONS

National power grid operations

🌦️
Climate

INMET

Meteorological time series

🛒
Public procurement

PNCP · Comprasnet

Contracts and tenders

and more

27 sources in total

A continuously expanding catalog, always with provenance

// FOR PEOPLE WHO LIVE ON RISK AND DECISIONS

From question to context, in seconds.

Risk, compliance, market intelligence and due diligence — four fronts where having all of Brazil in one graph changes the game.

🪪

KYC / KYB and onboarding

Validate a supplier or customer from their CNPJ: existence, ownership structure, tax standing and relationships — all before signing the contract, with an origin trail for every field.

CNPJ→ ownership→ tax standing→ decision
🕵️

Fraud prevention

Spot suspicious structures by relationship, not reputation: shared partners, chains of control and shared addresses linking seemingly independent entities.

ownership graph→ hidden links→ alert
📊

Market intelligence

Size a sector by municipality, map active competitors, follow a product's chain back to the raw material. The country's entire universe of companies as your analysis base — not a sample.

sector + region→ universe of CNPJs→ real TAM
🚨

Risk monitoring

Track a portfolio of counterparties over time: ownership changes, new tax restrictions, recorded sanctions. Data arriving in near real time, with provenance.

portfolio→ change detected→ review
⚠️

Indicators, not accusations. Risk signals point to relationships and situations recorded in public sources to guide your analysis — they are not a verdict. The decision always belongs to a person, based on evidence traceable back to its origin.

✔ LGPD by design

Public data, handled responsibly.

ZDZDataVerse consolidates data from official public sources. The scale is enormous; the exposure is controlled. We work with aggregates and capabilities — the platform never serves an individual person's record.

  • Official public sources only. 27 government sources, with no scraping of private or sensitive data outside a legal basis.
  • Provenance on every field. Origin, date and confidence — auditable down to the public document that generated the data.
  • Individuals protected. Masked CPF, recorded legal basis, no exposure of individual records.
  • A signal, not a sentence. Risk indicators guide human analysis; they never publicly accuse anyone by name.

Want all of Brazil working for your analysis?

Let's design a pilot around your risk, compliance or market intelligence cases. AI + People, traceable evidence, no lock-in.