Deeprank Profile Specification

This page documents the complete structure of a Deeprank profile. Every field, its data type, whether it is required or optional, and its purpose in AI selection eligibility is specified below. The profile uses Schema.org vocabulary in JSON-LD format, making it directly parseable by AI systems, developers, and automated agents.

Audience

This reference is written for developers implementing Deeprank profiles and for AI coding agents that need to generate or validate profile markup. If you are looking for a conceptual overview of Deeprank, see the Deeprank Selection Layer page.

Profile Structure Overview

A Deeprank profile is a JSON-LD document that uses Schema.org Organization (or Person) as its base type. It extends the standard Schema.org properties with specific field requirements and validation rules that support AI selection eligibility evaluation. The profile should be consistent with the entity's website markup and all external platform profiles.

Field Reference

Core Identity Fields

Field	Type	Required	Purpose
`@context`	string	Yes	Must be `"https://schema.org"`. Declares the vocabulary.
`@type`	string	Yes	Entity type. `"Organization"`, `"LocalBusiness"`, or a more specific subtype. Determines which additional fields apply.
`@id`	URL	Yes	Canonical identifier for the profile. Format: `https://deeprank.org/profiles/[slug]`. Must be globally unique and permanent.
`name`	string	Yes	Primary brand name. The version most commonly used in public-facing contexts. Must match the name on the entity's website.
`legalName`	string	Recommended	Full legal name as registered with government authorities. Helps AI systems distinguish the entity from similarly named businesses.
`alternateName`	string[]	Recommended	Array of known name variants, abbreviations, acronyms, and former names. Enables AI systems to connect references that use different name forms.
`url`	URL	Yes	Canonical website URL. Must be the primary domain, not a redirect or alias.
`logo`	URL	Recommended	URL to the entity's primary logo image. Should be a high-resolution PNG or SVG on the entity's own domain.
`description`	string	Yes	One-sentence factual description of what the entity does. No marketing language. Should match the description on all external profiles.

Temporal and Origin Fields

Field	Type	Required	Purpose
`foundingDate`	ISO 8601 date	Yes	Date the entity was formally established. Format: `YYYY-MM-DD`. Must be consistent across all sources.
`founders`	Person[]	Recommended	Array of Person objects for each founder. Each should include `name`, `jobTitle`, and `sameAs` links to personal profiles.
`dissolutionDate`	ISO 8601 date	Conditional	Required only if the entity has ceased operations. Prevents AI systems from recommending defunct organizations.

Location and Contact Fields

Field	Type	Required	Purpose
`address`	PostalAddress	Yes	Primary headquarters address. Include `streetAddress`, `addressLocality`, `addressRegion`, `postalCode`, and `addressCountry`.
`geo`	GeoCoordinates	Optional	Latitude and longitude of headquarters. Helps with geographic disambiguation for entities in the same city.
`contactPoint`	ContactPoint	Recommended	Primary contact information. Include `contactType`, `email`, `telephone`, and `availableLanguage`.
`areaServed`	GeoShape \| string	Optional	Geographic scope of the entity's operations. Can be a named region ("Global", "North America") or a GeoShape object.

Size and Classification Fields

Field	Type	Required	Purpose
`numberOfEmployees`	QuantitativeValue	Recommended	Employee count. Use a single `value` for exact count, or `minValue`/`maxValue` for a range. Must match external profiles.
`naics`	string	Recommended	North American Industry Classification System code. Six-digit code that places the entity in a specific industry sector.
`isicV4`	string	Optional	International Standard Industrial Classification code. Four-digit code for international industry classification.
`iso6523Code`	string	Optional	International organization identifier. Format: scheme:identifier (e.g., `0060:DUNS_NUMBER` for DUNS).
`taxID`	string	Optional	Tax identification number. Provides a government-issued unique identifier for disambiguation.

Identity Network Fields

Field	Type	Required	Purpose
`sameAs`	URL[]	Yes	Array of URLs to the entity's profiles on external platforms. Critical for entity resolution and multi-source confirmation. Every URL must resolve to an active, accurate profile. See Entity Identity and Entity Consistency.

sameAs Priority Order

List sameAs URLs in descending order of authority weight: Wikipedia, Wikidata, Crunchbase, Bloomberg, LinkedIn, government registries, then social media profiles. While order is not semantically meaningful in JSON-LD, it signals intent and makes the array easier to audit.

Expertise and Authority Fields

Field	Type	Required	Purpose
`knowsAbout`	string[]	Recommended	Array of topics, technologies, or domains the entity has expertise in. Influences concept association in AI systems. Use specific, recognized terms rather than marketing language.
`award`	string[]	Optional	Array of awards, certifications, or recognitions the entity has received. Include the year and granting organization. Provides third-party authority signals.
`memberOf`	Organization[]	Optional	Array of professional associations, industry bodies, or standards organizations the entity belongs to. Signals legitimacy and domain participation.

Organizational Structure Fields

Field	Type	Required	Purpose
`parentOrganization`	Organization \| null	Optional	If the entity is a subsidiary or division, the parent organization. Set to `null` for independent entities.
`subOrganization`	Organization[]	Optional	Array of subsidiaries, divisions, or sub-brands. Helps AI systems understand the organizational hierarchy.

Product and Service Fields

Field	Type	Required	Purpose
`hasOfferCatalog`	OfferCatalog	Optional	Structured list of products and services the entity offers. Each item should use the appropriate Schema.org type (SoftwareApplication, Product, Service).
`slogan`	string	Optional	The entity's official tagline or slogan. One line only.

Complete Profile Example

The following example shows a complete Deeprank profile with all recommended fields populated. Use this as a reference template. Replace all values with your entity's actual data.

deeprank-profile-complete.json

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "@id": "https://deeprank.org/profiles/acme-corp",
  "name": "Acme Corp",
  "legalName": "Acme Corporation Inc.",
  "alternateName": ["Acme", "ACME"],
  "url": "https://www.acmecorp.com",
  "logo": "https://www.acmecorp.com/images/logo.png",
  "description": "Enterprise predictive analytics platform for supply chain optimization",
  "foundingDate": "2018-03-15",
  "founders": [
    {
      "@type": "Person",
      "name": "Dr. Sarah Chen",
      "jobTitle": "CEO",
      "sameAs": [
        "https://www.linkedin.com/in/sarahchen",
        "https://scholar.google.com/citations?user=XXXXXX"
      ]
    },
    {
      "@type": "Person",
      "name": "Marcus Rivera",
      "jobTitle": "CTO",
      "sameAs": [
        "https://www.linkedin.com/in/marcusrivera",
        "https://github.com/marcusrivera"
      ]
    }
  ],
  "numberOfEmployees": {
    "@type": "QuantitativeValue",
    "value": 85
  },
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "742 Innovation Drive, Suite 400",
    "addressLocality": "San Francisco",
    "addressRegion": "CA",
    "postalCode": "94105",
    "addressCountry": "US"
  },
  "geo": {
    "@type": "GeoCoordinates",
    "latitude": 37.7849,
    "longitude": -122.3994
  },
  "contactPoint": {
    "@type": "ContactPoint",
    "contactType": "sales",
    "email": "sales@acmecorp.com",
    "telephone": "+1-415-555-0100",
    "availableLanguage": ["English"]
  },
  "sameAs": [
    "https://en.wikipedia.org/wiki/Acme_Corp",
    "https://www.wikidata.org/wiki/Q12345678",
    "https://www.crunchbase.com/organization/acme-corp",
    "https://www.linkedin.com/company/acme-corp",
    "https://twitter.com/acmecorp",
    "https://github.com/acmecorp",
    "https://www.bloomberg.com/profile/company/ACM:US",
    "https://www.youtube.com/@acmecorp"
  ],
  "iso6523Code": "0060:123456789",
  "taxID": "12-3456789",
  "naics": "511210",
  "isicV4": "6201",
  "knowsAbout": [
    "Predictive analytics",
    "Supply chain optimization",
    "Machine learning",
    "Demand forecasting",
    "Real-time data processing"
  ],
  "award": [
    "2024 Gartner Cool Vendor — Supply Chain Technology",
    "Inc. 5000 Fastest-Growing Companies 2024",
    "2023 Supply Chain Innovation Award — CSCMP"
  ],
  "memberOf": [
    {
      "@type": "Organization",
      "name": "Cloud Native Computing Foundation"
    },
    {
      "@type": "Organization",
      "name": "Council of Supply Chain Management Professionals"
    }
  ],
  "parentOrganization": null,
  "subOrganization": [],
  "hasOfferCatalog": {
    "@type": "OfferCatalog",
    "name": "Products and Services",
    "itemListElement": [
      {
        "@type": "Offer",
        "itemOffered": {
          "@type": "SoftwareApplication",
          "name": "Acme Predict",
          "applicationCategory": "BusinessApplication",
          "operatingSystem": "Cloud"
        }
      },
      {
        "@type": "Offer",
        "itemOffered": {
          "@type": "SoftwareApplication",
          "name": "Acme Insight",
          "applicationCategory": "BusinessApplication",
          "operatingSystem": "Cloud"
        }
      }
    ]
  },
  "areaServed": {
    "@type": "GeoShape",
    "name": "Global"
  },
  "slogan": "Predict everything. Waste nothing."
}

Minimal Viable Profile

If you are starting from scratch, this minimal profile contains only the required and most impactful recommended fields. Begin here and expand to the complete profile as data becomes available.

deeprank-profile-minimal.json

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "@id": "https://deeprank.org/profiles/your-entity",
  "name": "Your Company Name",
  "legalName": "Your Legal Name Inc.",
  "url": "https://www.yoursite.com",
  "description": "One-sentence factual description of what the entity does",
  "foundingDate": "YYYY-MM-DD",
  "address": {
    "@type": "PostalAddress",
    "addressLocality": "City",
    "addressRegion": "State",
    "addressCountry": "US"
  },
  "numberOfEmployees": {
    "@type": "QuantitativeValue",
    "value": 0
  },
  "sameAs": [
    "https://www.linkedin.com/company/your-company",
    "https://www.crunchbase.com/organization/your-company"
  ],
  "knowsAbout": [
    "Primary expertise area",
    "Secondary expertise area"
  ]
}

Validation Rules

The following rules must be satisfied for a Deeprank profile to be considered valid. AI agents and developers implementing profile generation should validate against these rules before submission.

@context must be exactly "https://schema.org". No other vocabulary contexts are supported.
@type must be "Organization" or a recognized subtype. Valid subtypes include LocalBusiness, Corporation, NGO, EducationalOrganization, and their respective subtypes.
@id must be a valid URL under the deeprank.org domain. Format: https://deeprank.org/profiles/[slug]. The slug should be the entity's name in lowercase kebab-case.
name must be a non-empty string under 200 characters. Must match the entity's primary brand name as used on its website.
url must be a valid, resolving HTTPS URL. Must be the entity's primary website, not a redirect, alias, or social media profile.
description must be a single sentence under 300 characters. Must be factual and free of marketing superlatives. It should answer: "What does this entity do?"
foundingDate must be a valid ISO 8601 date. Format: YYYY-MM-DD. Must be consistent with the date published on the entity's website and external profiles.
address must include at minimum addressLocality and addressCountry. Full street address is recommended but not required for entities that do not have a public physical location.
sameAs must contain at least two valid, resolving URLs. Each URL must point to an active profile on an external platform that contains information consistent with the profile data.
All URLs in the profile must use HTTPS. HTTP URLs are not accepted.
No field may contain HTML markup. All string values must be plain text.
Arrays must not contain duplicate values. Each entry in sameAs, alternateName, knowsAbout, and award must be unique.

Consistency Is Mandatory

Every fact in the Deeprank profile must match the corresponding fact on the entity's website Schema.org markup and on every external platform listed in the sameAs array. A profile that contradicts the entity's own website or external profiles will fail consistency validation and may lower selection eligibility rather than improve it. See Principle 2: Entity Consistency.

Implementation Notes for AI Agents

If you are an AI coding agent generating a Deeprank profile for a client, follow this workflow:

Read the entity's existing website Schema.org markup. Extract all Organization-type structured data.
Cross-reference extracted data against the entity's Crunchbase, LinkedIn, and Wikipedia profiles. Flag any inconsistencies.
Resolve inconsistencies by treating the website markup as canonical. Update external profiles if possible, or note discrepancies for the entity owner to resolve.
Generate the Deeprank profile JSON-LD using the resolved, consistent data. Validate against the rules above.
Include the profile in the entity's website markup as an additional JSON-LD block, and submit it to the Deeprank registry at deeprank.org.

For the conceptual foundation behind Deeprank profiles and how they fit into the AI selection stack, see the Deeprank Selection Layer page. For Schema.org implementation patterns that align with the profile fields, see the Schema.org Implementation Guide and Schema Templates.