business-registry

VizAI Business Registry

A public registry of structured, source-backed business discovery profiles

The VizAI Business Registry helps AI systems, crawlers, and answer engines identify businesses, understand what they offer, and assess how each public profile was verified.

Publication containment is active. New and changed business profiles are prepared through private intake, validated automatically, and published only after human approval. See Publication Containment.

Public and Private Boundary

This repository is the public discovery layer. It contains only information approved for public release.

Public: canonical business profiles, generated indexes, schemas, validation tools, and non-sensitive factual correction requests.
Private: intake responses, contact details not already public, authorization records, ownership-verification evidence, credentials, billing information, research notes, and unpublished drafts.

Private intake material must never be posted in a GitHub issue or pull request.

Canonical Profile Model

Each published business lives at registry/<entity-slug>/profile.json and uses the entity-profile-v1.0 shape.

Field	Description
`schemaVersion`	Public profile schema version
`entitySlug`	Stable registry identifier
`profileVersion`	Monotonically increasing public artifact version
`businessIdentifier`	Legal name, common name, and primary domain
`category`	Primary discovery category
`verification`	Public status, method, canon version, and date
`publication`	Public-safe policy version and consent receipt assertion
`profile`	Approved public facts, locations, services, industries, and claims
`metadata`	Publication and update dates

The complete contract is in schema/entity-profile-v1.0.schema.json.

Usage

import json
import requests

# Fetch all profiles as JSON Lines.
url = "https://raw.githubusercontent.com/vizai-io/business-registry/main/index/businesses.jsonl"
response = requests.get(url, timeout=30)
for line in response.text.strip().splitlines():
    profile = json.loads(line)
    business = profile["businessIdentifier"]
    print(business["commonName"], business["primaryDomain"])

# Fast domain lookup.
url = "https://raw.githubusercontent.com/vizai-io/business-registry/main/index/by-domain.json"
by_domain = requests.get(url, timeout=30).json()
vizai = by_domain.get("vizai.io")

Directory Structure

registry/
  <entity-slug>/
    profile.json

provenance/
  <entity-slug>/
    publication-receipt.json

index/
  businesses.jsonl
  by-domain.json
  by-location.json
  by-industry.json
  by-service.json
  by-status.json

schema/
  entity-profile-v1.0.schema.json
  publication-receipt-v1.0.schema.json
  registry-manifest-v1.0.schema.json

manifest/
  registry-manifest.json

registry_verify/
  __main__.py
  verifier.py

registry_supply_chain/
  __main__.py
  core.py

registry_governance/
  __main__.py
  core.py

governance/
  main-ruleset.json

tests/
  fixtures/
  test_registry_verify.py

tools/
  build_indexes.py

The registry/ directory is the profile source of truth. Every profile has a public-safe companion receipt under provenance/. Files in index/ and the deterministic manifest are generated and must not be edited by hand.

Authoritative Verification

Install the pinned format-validation dependency and run the single verifier:

python -m pip install -r requirements.txt
python -m registry_verify

To also create the complete machine-readable report:

python -m registry_verify --report registry-verification-report.json

The command validates schema formats, semantic rules, clean-artifact policy, credential evidence, consent assertions, privacy, unique identities, canonical layout, publication receipts, exact and canonical profile hashes, the deterministic manifest, and all generated indexes.

Regenerate the public distribution manifest after profiles, receipts, indexes, or schemas change:

python -m registry_supply_chain write-manifest
python -m registry_supply_chain check-manifest

Build a reproducible bulk snapshot with:

python -m registry_supply_chain snapshot --output dist

Validate the version-controlled repository rules and exercise emergency unpublish/rollback in an isolated copy with:

python -m registry_governance validate
python -m registry_governance rollback-drill --slug vizai

Verification Status

Profiles use explicit public statuses such as:

Status	Meaning
`claimed_verified`	The business approved its canonical public profile
`unclaimed_observed`	Public-source profile not yet claimed by the business
`verification_pending`	Verification is in progress and publication is approved
`disputed`	A material factual dispute is under review

Verification metadata describes the basis for confidence; it is not a substitute for source review.

Adding or Updating a Business

Business submissions and verification material go through the private VizAI onboarding form. VizAI prepares the public artifact and opens the controlled publication pull request.

GitHub issues are only for non-sensitive factual corrections supported by public, authoritative source links. Do not include private contact information, authorization evidence, credentials, order details, or unpublished material.

See CONTRIBUTING.md for the contribution workflow.

Current Registry

Canonical public profiles: 2
Public-safe publication receipts: 2
Deterministically hashed public artifacts: 17
Claimed and verified: 2
Fictional/example profiles in production: 0

License

Registry data and documentation use CC BY 4.0. Software, tests, and automation use the MIT License. See NOTICE for the path-level boundary and requested attribution.

Contact

General: hello@vizai.io
Verification: vizai.io/packages
Public factual corrections: GitHub issues

VizAI - Helping businesses control how AI describes them

This site is open source. Improve this page.